Incomplete preparation and unavailable supplies, beds, and equipment are operational hypotheses to investigate when a procedure is canceled, without assuming how often they occur or whether they were avoidable. The proposed method organizes these dependencies, preserves clinical decision-making, and allows teams to learn from failures without turning maintenance or the patient into an automatic explanation.

Define what counts as a cancellation

Before discussing targets, establish the point at which the schedule becomes the reference for the metric. One local option is to take a snapshot of the schedule at the close of the previous day. Procedures removed after that point count as cancellations from that schedule; earlier changes are tracked in a separate measure. This is a proposed management convention that must be documented and applied consistently.

Also record delays, rescheduling, and room changes. For each cancellation, separate the primary reason, contributing factors, and the moment the open issue was identified. Unavailable equipment may be the immediate reason, while the lack of a reservation or of communication was the factor that made the impact unavoidable. Classification requires multidisciplinary review and allows for causes that remain undetermined.

Schedule around each procedure's dependencies

RDC 63/2011, an ANVISA regulation on good operating practices for health services, requires equipment, supplies, consumables, and medications to be available in line with demand and provides for health technology management. This foundation supports the need to organize requirements, without setting a single confirmation window for all surgeries. [1]

For each procedure, link the team, room, estimated duration, instruments, specific supplies, required equipment, and post-surgical destination. Distinguish between a resource that is requested, reserved, and actually available. The expected bed must be confirmed with capacity management at the right time, because an intended discharge is not the same as a vacated bed. Emergencies and clinical changes require an explicit reassessment of the schedule.

Check preparation and communication earlier

Define checkpoints proportional to the procedure: one review when the case enters the schedule and another before operational confirmation. These windows are choices made by the service, not universal deadlines. The clinical team defines the applicable clinical requirements; coordination tracks whether they were verified, who owns each open issue, and when the decision must be made. Avoid recording only a ready mark with no author or time.

Communication with the patient should use approved instructions and a channel for answering questions. Confirm receipt and understanding when relevant, taking accessibility and travel barriers into account. If a clinical condition is not met, the decision rests with licensed professionals. Operational consulting does not prescribe preparation, medication holds, or anesthesia criteria. Its role is to keep a known question from staying invisible until the patient arrives in the room.

Treat supplies as resources with deadlines and evidence

Physical stock alone does not demonstrate availability for a case. The proposal is to track quantity, compatibility, integrity, expiration, and release status, in line with the relevant technical procedures. For items that require reprocessing, coordinate the sterile processing department (SPD) with the schedule. For consigned or third-party supplies, record the expected delivery, inspection, and the person responsible for resolving discrepancies.

When two rooms depend on the same instrument set, a double reservation creates fictitious availability. Use a link between case and resource, with an overlap alert and a turnaround time defined by the actual process. If the supply cannot be confirmed, the open issue must reach someone with authority to decide. Last-minute purchasing or borrowing cannot be the only contingency described.

Integrate maintenance and scheduling without assuming release

RDC 509/2021, ANVISA's regulation on health technology management, requires a management plan, systematic records, defined responsibilities, and risk monitoring for the technologies it covers. Equipment must be managed throughout its life cycle, not only when a failure occurs. [2]

As an operational design, connect the clinical engineering inventory to the schedule by equipment identifier. Distinguish between available, reserved, under evaluation, under maintenance, and awaiting technical release. Closing an administrative work order should not automatically be interpreted as clearance for use. The evidence required to return equipment to service depends on the intervention, the technical procedures, and the instructions that apply to that equipment.

Cross-check scheduled maintenance against the procedures that depend on that resource before closing the schedule. In a contingency, confirm whether an alternative has compatible capacity, accessories, and qualified staff. A device located in another unit may already be committed to other patient care. The decision to reassign it must consider that impact and have an identified owner, especially outside business hours.

Preserve the purpose of the surgical safety checklist

The Brazilian national protocol describes the checklist at three points: before induction of anesthesia, before skin incision, and before the patient leaves the operating room. It also addresses equipment condition and team communication. Advance verification of resources should complement these safeguards while preserving the check at the point of care. [3]

Implementation requires team participation and feedback on practice. The WHO manual highlights local leadership, education, multidisciplinary involvement, and adaptation to context. [4] In the approach proposed here, the advance review identifies logistical obstacles; the surgical checklist keeps its clinical purpose. Filling in a screen ahead of time does not prove that the safety conversation took place in the room.

Decide the contingency and record the effect on the patient

Each critical open issue should have a path: fix it by the agreed deadline, replace it with a technically validated alternative, reorder the schedule, or escalate the postponement to the responsible decision-maker. Record who decides, who will be informed, and what information goes into the shift handoff. An emergency may call for a different priority, but that choice does not justify ignoring safety or availability requirements.

Worked example: in one month, 20 procedures were canceled after the schedule closed. Each event was assigned a primary reason: eight for supplies, five for preparation, four for equipment, and three for beds. The breakdown suggests where to start the investigation, without demonstrating avoidability or the effect of an intervention.

Test the workflow on a defined schedule

Start with one specialty or set of rooms and keep an open-issues log with owner, deadline, evidence, and decision. Bring together scheduling, nursing, the medical team, supplies, bed management, and clinical engineering to examine a few complete cases. Look for the first opportunity when information could have changed the outcome, without assuming the unit where the problem surfaced is to blame.

Compare periods using the same schedule definition and report volume, procedure mix, and emergencies. Lowering the rate by removing difficult cases before the cutoff may simply shift the problem. For that reason, also track wait time until rescheduling, changes made before the schedule closes, and repeat cancellations. The next verifiable step is to show that every critical open issue had a recorded decision before the time planned for the procedure.

Implementation checklist

• Document the schedule cutoff point and the definition of cancellation.
• Link critical requirements and resources to each scheduled procedure.
• Record the owner, deadline, and evidence of resolution for open issues.
• Check instrument reservations and the release status of supplies.
• Cross-check the schedule against maintenance and equipment technical availability.
• Validate alternatives and contingency contacts for every shift.
• Review cancellations with contributing factors and patient communication.

Tracking indicators

Cancellation after schedule close

100 × procedures canceled after the defined cutoff / procedures in the schedule snapshot at that cutoff, in the period.. Report counts, the cutoff, and exclusions. Separate emergencies added later and track changes made before the cutoff.

Critical open issues resolved on time

100 × critical open issues with documented resolution by the agreed deadline / critical open issues due in the period.. Also track the number of cases with open issues; late entries can distort the result.

Impact of equipment unavailability

Number of canceled procedures with equipment as the primary reason and number with equipment as a contributing factor, in separate series.. Do not add the two series without deduplication. Confirm the link to the equipment and review the cause.

Wait time to rescheduling

Median and 90th percentile of days between cancellation and completion of the procedure, by cancellation cohort.. Also report how many cases remain unperformed; silently excluding them flatters the result.

Process map: from scheduling to cancellation review

1. Define requirements: The clinical team and scheduling link needs to the procedure.
2. Confirm resources: Supplies, SPD, beds, and clinical engineering record evidence and constraints.
3. Resolve open issues: The designated owner fixes or escalates before the operational deadline.
4. Decide contingency: Coordination and responsible professionals evaluate an alternative, reordering, or postponement; they return to resource confirmation if anything changes.
5. Perform clinical checks: The team applies the safe surgery protocol at the designated points.
6. Review outcome: Management reviews completion, delay, or cancellation and feeds the findings back into planning.

Primary reasons for 20 cancellations in one month

Each cancellation is assigned one primary reason. Supplies: 8/20; preparation: 5/20; equipment: 4/20; beds: 3/20. Counts for discussing prioritization, without inferring national causes.

• Supplies: 8 canceled procedures
• Preparation: 5 canceled procedures
• Equipment: 4 canceled procedures
• Beds: 3 canceled procedures

Example to illustrate the method.

Frequently asked questions

What is an acceptable surgical cancellation rate?

This article does not set a universal target. Use a stable definition, stratify the schedule mix, and analyze safety, access, and causes before setting local goals.

Does the surgical safety checklist replace advance confirmation?

No. The national protocol calls for checks at specific points of care. Advance confirmation organizes logistical dependencies and must coexist with those safeguards. [3]

Is every equipment-related cancellation avoidable?

That cannot be concluded from the recorded reason alone. Investigate the failure, technical condition, maintenance, reservation, alternatives, and timing of communication before classifying avoidability.

Sources

1. RDC 63/2011: boas práticas de funcionamento para serviços de saúde
2. RDC 509/2021: gerenciamento de tecnologias em saúde
3. Protocolo para cirurgia segura
4. Implementation manual WHO surgical safety checklist 2009

This guide offers a simple way to measure the real capacity of a diagnostic imaging service or an outpatient clinic that schedules by room. The core idea is to distinguish four quantities that tend to get blended together: available capacity, scheduled capacity, delivered capacity, and the step that limits the flow. The recommendations are operational and need to be adapted to each service.

Separate four numbers before discussing expansion

Available capacity is the time the room can actually see patients, after subtracting preventive maintenance, cleaning, shift changes, and scheduled breaks. Scheduled capacity is the portion of that time committed to booked exams or held for add-ons. Delivered capacity is the time actually taken up by completed exams. The gaps between these numbers show where the schedule loses hours, and each loss has a different owner and cause.

The fourth element is the bottleneck, the step with the lowest capacity in the end-to-end flow. It may be the room, but it may also be registration, patient prep, the technologists available in that time slot, or report sign-off. Expanding a step that does not limit the flow rarely shortens the patient's wait. IHI frames flow improvement as shaping demand, matching capacity to demand, and redesigning the system, using data to predict demand patterns and reduce artificial variation in scheduling. [2]

National benchmarks also have a limited role. In Brazil's public health system (SUS), planning criteria and parameters are quantitative references for estimating needs and guiding managers; they are indicative, not mandatory, and can be adapted to local conditions. [1] They help size the network, but they do not tell you how much a specific room delivers today. That answer comes from the service's own records.

Define the available hours for each room

Start with one profile per room and modality: operating hours, days of service, scheduled breaks, maintenance windows, cleaning or turnover time between patients, and periods with no technologist on the schedule. A room open twelve hours but staffed for only eight has eight hours of available capacity for that exam type. Record the rule and its effective date, because staffing schedules and maintenance contracts change over the year.

Next, estimate the duration of each exam type, counting room time from positioning to release, not just acquisition time. A plain radiograph, a Doppler ultrasound, and a contrast CT occupy the room for very different lengths of time. Use medians from recent records, by exam type and by shift, rather than a single value carried over from when the system was implemented. If the system does not capture start and end times, two weeks of manual sampling is enough to show the order of magnitude.

• Operating hours: what the contract or staffing schedule says.
• Available hours: operating hours minus maintenance, cleaning, breaks, and staffing gaps.
• Theoretical slots: available hours divided by the median duration of each exam type, following the planned case mix.

Compare the time held for add-ons with what actually happens

Any imaging schedule that serves the emergency department or inpatient units needs room for urgent cases. The problem shows up when the planned hold does not match reality. Take, as an example, a CT schedule that holds 20% of available hours for add-ons and urgent cases, while the records show urgent cases taking up 30%. The difference does not go away: it turns into elective exams that run late, get rescheduled, or are done on overtime.

The answer is not necessarily to raise the hold to the highest value observed. Look at the distribution by day of the week and time of day, because urgent cases tend to cluster in predictable periods. A larger hold at certain times and a smaller one at others can absorb the demand without blocking slots that would sit empty. Decisions about the clinical priority of each order remain with the responsible clinicians; the schedule design simply creates space that fits the observed pattern.

Treat no-shows and cancellations as measurable losses

A no-show is a direct loss of capacity when the slot is not backfilled. A study of more than 3.3 million outpatient radiology appointments at an academic institution in the United States, from 2000 to 2015, recorded 8.0% cancellations and 2.6% no-shows, with no-shows associated with factors such as income, travel distance, and weather. [3] Those percentages belong to that setting and should not be used as a benchmark for a Brazilian service.

The useful part is the method: measure no-shows and late cancellations separately, by modality, time slot, and booking lead time, and check how many of those slots were filled by another patient. Active confirmation, a same-day call list, and prep instructions sent well in advance are hypotheses to test. Before adopting overbooking, assess its effect on delays and on the experience of the patients who did show up.

Include patient prep and reporting in the flow, not just the room

An exam that cannot be performed because the patient arrived without fasting, without a required lab result, or without authorization uses up room time and produces nothing. Record the reason for every exam not performed, using fixed categories. If incomplete prep keeps recurring, the fix lies in patient instructions, advance checks, or communication with the ordering unit, not in the number of rooms.

From the ordering clinician's point of view, the service is done when the report is available. A study in an emergency department in the Netherlands broke total radiology department time into its components and, after Lean interventions that included a dedicated radiologist during peak hours, found the median total time fell from 70 to 36 minutes and the median report time from 37 to 11 minutes, with exam time essentially unchanged. [4] The takeaway for managers is that the delay was outside the room. Measuring scanner utilization alone would have hidden that step.

Worked example: where 50 hours of room time went

Picture a CT room with 60 operating hours per week and 50 available hours after maintenance, cleaning, and breaks. The schedule planned 40 hours of elective exams and 10 hours, or 20%, for add-ons and urgent cases. In the week observed, urgent cases and add-ons took up 15 hours, or 30% of available time: the 10 hours held, 4 hours taken from elective exams that were then rescheduled, and 1 hour of no-show time that was backfilled.

Of the 40 elective hours, 4 were displaced by urgent cases and 4 were no-shows, 3 of which sat idle. Another 3 hours of room time were lost to unprepared patients and 1 hour to longer than expected gaps between exams. That left 28 hours of completed elective exams. Adding 28 and 15, the room performed exams in 43 of its 50 hours, 86% of available time. The remaining 7 hours do not call for a new room; they call for rules on backfilling, patient prep, and turnaround between exams.

The chart shows this breakdown. The numbers illustrate the math and are not an industry benchmark. In practice, the same analysis should be repeated over several weeks before drawing any conclusion, because one atypical week skews the picture.

Build a short dictionary of capacity metrics

Each metric needs a formula, a source, a frequency, and an owner, so that two people reach the same number from the same records. The list below is a minimal proposal, to be adjusted to the systems you have. The full formulas and notes on interpretation are in the metrics section of this article.

• Delivered room utilization: hours with a completed exam divided by available hours. Source: scheduling system and start and end time records. Frequency: weekly. Owner: imaging service manager.
• Utilization by urgent cases and add-ons: urgent and add-on hours divided by available hours, compared with the planned hold. Source: schedule with order origin flagged. Frequency: weekly, by time slot. Owner: scheduling lead.
• No-show rate: no-shows divided by confirmed scheduled exams. Source: schedule and front desk. Frequency: weekly, by modality. Owner: central scheduling.
• Prep-related loss: room hours lost to exams not performed because of incomplete prep. Source: reason recorded in the room. Frequency: monthly. Owner: nursing or the technologist in charge of the shift.
• Time to report: median time from end of exam to report sign-off, by modality and priority. Source: reporting system. Frequency: weekly. Owner: the service's medical director.

Test a scheduling change before buying a room

Pick one modality and one well-defined change: redistribute the urgent case hold by time slot, set up a call list for slots freed by no-shows, or check patient prep two days before the exam. Define the test period, the baseline, the metrics to track, and a balancing measure, such as wait time for urgent cases, to make sure an improvement in one place does not make things worse in another.

At the end, compare delivered hours, time to report, and wait time per exam against the baseline. If demand still exceeds delivered capacity even after losses come down, the case for expansion gets stronger, because it shows how much capacity is missing and at which step. This is an operational recommendation, not a promise of results; the outcome depends on the service, its demand, and local constraints.

Implementation checklist

• Record operating hours and truly available hours for each room.
• Measure median room time by exam type and shift.
• Compare the hold planned for add-ons with actual use by time slot.
• Track no-shows, late cancellations, and backfilled slots separately.
• Classify exams not performed because of prep using fixed categories.
• Measure the time from end of exam to report sign-off.
• Approve the formula, source, frequency, and owner for each metric.
• Test a scheduling change against a baseline before proposing a new room.

Tracking indicators

Delivered room utilization

100 × room hours with a completed exam / available room hours in the period, net of maintenance, cleaning, and scheduled breaks.. High utilization with a growing backlog suggests a real constraint; low utilization with a backlog points to schedule losses or a bottleneck outside the room. There is no universal target.

Utilization by urgent cases and add-ons

100 × hours used by urgent cases and add-ons / available hours, shown next to the planned hold for the same period and time slot.. The gap between planned and actual shows how much elective time was displaced. Flag the order origin at the time of scheduling.

No-show rate for scheduled exams

100 × patients who did not show up / exams scheduled in the period, by modality; cancellations with advance notice tracked as a separate series.. Track it alongside the share of backfilled slots, which shows how much of the no-show volume became an actual loss of capacity.

Time to report

Median time, in hours or minutes, from end of exam to report sign-off, by modality and priority.. Also show a high percentile, such as the 90th, to surface slow cases. Changes to the sign-off workflow will shift the series.

Process map: from order to report, with points where capacity is lost

1. Receive the order: Central scheduling classifies the order's origin, priority, and exam type.
2. Schedule and confirm: The schedule assigns an elective slot or add-on hold; confirmation cuts no-shows that are never backfilled.
3. Check patient prep: The team confirms fasting, prior tests, and authorization before the exam day; open items go back to the ordering unit.
4. Perform the exam: The technologist records start, end, and the reason for any exam not performed.
5. Issue the report: The radiologist signs off the report; time to sign-off is measured by modality and priority.
6. Review capacity: Management compares available, scheduled, and delivered capacity and decides on the next change.

How 50 available hours in a CT room were used

Example of one week: elective exams completed, 28 h; urgent cases and add-ons, 15 h (30% of the 50 h, versus 20% planned); no-shows not backfilled, 3 h; room idle because of incomplete prep, 3 h; longer than expected gaps between exams, 1 h. Total: 50 h.

• Elective exams completed: 28 room hours per week
• Urgent cases and add-ons: 15 room hours per week
• No-shows not backfilled: 3 room hours per week
• Room idle due to incomplete prep: 3 room hours per week
• Delays between exams: 1 room hours per week

Example to illustrate the method.

Frequently asked questions

What is the ideal utilization rate for an imaging room?

This article does not set a universal target. The right level depends on the exam mix, urgent demand, and the slack needed to absorb variation. Set the target locally, together with wait time metrics.

Is overbooking worth it to make up for no-shows?

It can be a hypothesis, but only after measuring no-shows by modality and time slot. Overbooking without data increases delays for the patients who do show up. Test it over a defined period and track wait times.

Why include report time in a capacity analysis?

Because the ordering clinician can only act on the result. If reports are slow, adding rooms increases the number of exams waiting to be read without shortening the time to a decision. [4]

Sources

1. Portaria de Consolidação nº 1, de 28 de setembro de 2017, arts. 102 a 106 (origem: Portaria GM/MS 1.631/2015)
2. Achieving Hospital-wide Patient Flow (Second Edition)
3. Environmental Factors Predictive of No-Show Visits in Radiology: Observations of Three Million Outpatient Imaging Visits Over 16 Years
4. Lean-driven interventions, including a dedicated radiologist, improve diagnostic imaging turnaround time and radiology report time at the emergency department

Absence coverage, uneven assignments, and burnout come together in daily operations, but they have no single solution. This article proposes organizing the work on three levels: determining staffing levels, building the schedule, and assigning care on each shift. The numerical examples illustrate the method. The method supports management decisions and requires judgment from the responsible nurse, local conditions, and the applicable professional and labor obligations.

Which Cofen reference to use for staffing levels

In the official texts consulted on October 26, 2026, Cofen Resolution 743/2024 (Cofen is Brazil's Federal Nursing Council) expressly revokes Resolution 543/2017 and points to Normative Opinion 1/2024 as the reference on the subject. A spreadsheet based only on the old resolution therefore needs its basis reviewed. Resolution 743/2024 is the act of revocation; the technical parameters are in the opinion. [1][2]

Opinion 1/2024 takes into account the characteristics of the service, the team, and the patients. It sets a minimum technical safety index (IST) of 15%, a minimum addition of 5% to the overall headcount for continuing education, including coverage related to turnover, and separate staffing for management activities. It also distinguishes methods by type of service. [2] These components must appear in the calculation worksheet with identifiable assumptions. Adding a generic margin to the total headcount is not enough, nor is assuming the minimum absorbs any local absence pattern.

Separate required headcount, schedule, and shift assignment

Determining staffing levels means estimating the headcount needed to sustain the service. The schedule distributes that headcount's availability across the calendar. Shift assignment matches the staff present to specific patients, activities, and responsibilities. When these decisions are mixed together, a well-organized schedule can hide a structural shortfall, while an apparently adequate headcount can be poorly distributed between high- and low-demand hours.

As a complementary methodological reference, the World Health Organization's WISN method starts from workload and time standards for each activity component. [3] This logic helps challenge the use of bed count alone as a measure of demand, without replacing the applicable Brazilian parameters. In the local analysis, make explicit which activities are included in the estimate, which depend on a specific role or competency, and how the volume data were obtained. An estimate can only be discussed when its assumptions are visible.

Measure demand and capacity without double counting

Collect data by unit, day, shift, and professional role. Record patient volume, the relevant patient classification, admissions, transfers, and activities that concentrate effort at certain times. Separate nominal availability from actual presence. Staff on vacation, on leave, or in training away from patient care do not provide the same hours for direct coverage during that period. At the same time, do not deduct again an unavailability already accounted for in the calculation method.

For inpatient units, the opinion uses patient classification and hours of care, with references of 4, 6, 10, 10, and 18 hours per patient for the minimal, intermediate, high-dependency, step-down, and intensive care categories. [2] These are 24-hour references, not individual quotas per shift. In operations, distribute the need across the day based on observation and the care profile. Counting inpatients without distinguishing their needs can produce equality that is only apparent.

Turn the assessment into scheduling rules

Before choosing a tool, define verifiable criteria: coverage by role, required competencies, contracted hours, work restrictions, and applicable rest rules. Distinguish mandatory requirements from negotiable preferences, such as certain time-off requests. Leadership should be able to explain why a proposal was accepted or rejected. Exceptions need a justification and an owner; they cannot disappear into a series of informal swaps.

An automated schedule should surface conflicts and preserve professional review. Test situations such as leave starting mid-month, the departure of a staff member with a specific competency, and a change in the unit's operating hours. Also consider whether reassignment takes capacity away from another unit. The operational gain may be less time spent building the schedule, but that only helps if actual coverage and workload distribution are monitored. A quick solution that generates more corrections later does not fix the process.

Cover call-outs and investigate absence patterns

Set up a workflow for unplanned absences: notification through an official channel, confirmation of impact, evaluation of alternatives, and a recorded decision. Leadership checks attendance, competencies, and demand before calling in coverage or reassigning staff. Senior management needs to be informed when the alternatives do not restore the required capacity. Repeated overtime as the default response can mask a staffing need or a persistent organizational problem.

When monitoring absences, establish a definition of absenteeism and keep vacations, open positions, and leaves of absence separate, depending on the purpose of the metric. Analyze aggregate patterns by unit and shift, with restricted access to individual information. Do not use health data for a punitive ranking. Unstable schedules, working conditions, leadership, and turnover are hypotheses to investigate with staff and HR; a coincidence between call-outs and overloaded shifts does not, on its own, prove a causal relationship or identify individual responsibility.

What the example chart can show

The chart shows an example of four shifts in one unit over a week. Each shift received its own estimate of need and a tally of hours actually available for patient care, using the same definition of time. In scenario A, 60 hours were needed and 60 provided; in B, 60 and 48; in C, 72 and 60; in D, 48 and 48. The deficit is the positive difference between need and availability: zero, twelve, twelve, and zero hours.

These values were invented solely to teach how to read the chart and do not result from a complete normative calculation. Two shifts with a twelve-hour deficit may require different responses depending on the role and competency that is missing. In addition, an adequate aggregate balance can hide an uncovered time window. That is why the dashboard must allow users to examine the hours and roles involved. Adding up RN and nursing technician hours indiscriminately can hide a gap that reassignment across roles cannot close.

Test the change and watch for unintended effects

Start with one unit, with participation from leadership, shift staff, HR, and senior management. Document the baseline, the demand method, and the coverage rules. Run a proposed schedule in parallel with the usual review before replacing the current process. Compare the conflicts found, the actual feasibility of coverage, and the time needed to correct the proposal. Keep the versions so you can understand why each decision was made.

Track coverage by role, overtime, reassignments, and changes after publication. Also watch for reports of overload and pending care activities, without treating a single fluctuation as a definitive result. The first useful deliverable is a reviewable calculation worksheet and an absence workflow that someone can actually execute. If the analysis shows a structural deficit, the decision must reach whoever can adjust resources and capacity. Organizing the schedule helps use the available headcount; hiring or reorganizing the service requires its own management decision.

Implementation checklist

• Review the regulatory basis of the spreadsheet and record the date Cofen was consulted.
• Document the demand method and calculation worksheet by service and role.
• Check additions, unavailability, and separately staffed activities individually.
• Define mandatory rules, competencies, and preferences before building the schedule.
• Confirm attendance and capacity in the home unit before reassigning anyone.
• Assign a channel, an owner, and an escalation path for unplanned absences.
• Compare actual coverage, schedule changes, and overtime during the pilot.
• Escalate persistent deficits to senior management with evidence of demand and capacity.

Tracking indicators

Actual coverage by role

Hours actually available for patient care / estimated required hours, in the same unit, role, and shift × 100.. Do not offset an RN deficit with a surplus of nursing technicians. The quality of the demand estimate limits the metric.

Shifts with a deficit

Shifts with at least one interval below the defined requirement / shifts observed × 100.. Report role, duration, and coverage rule; the monthly total of hours can hide short-term gaps.

Unplanned absence

Hours of unplanned absence / scheduled hours originally planned for patient care in the period × 100.. An operational convention of this article; define which events are included and freeze the denominator so adjustments do not erase call-outs.

Overtime intensity

Overtime hours actually worked / total hours actually worked in the period × 100.. Distinguish emergency coverage from planned activity and reconcile records with HR.

Changes after publication

Published shifts with at least one change / published shifts in the period × 100.. Classify the reason and distinguish staff-requested swaps from changes needed to cover a deficit.

Process map: from demand to schedule review

1. Characterize demand: Responsible nurse: care profile, activities, and variation by time of day.
2. Determine staffing levels: Nursing and management: calculation worksheet, roles, and required resources.
3. Assess availability: Leadership and HR: work hours, vacations, leaves, and competencies.
4. Plan and review schedule: Leadership: validate coverage, rules, and conflicts before publication.
5. Assign the shift: Shift nurse: match current demand to the capacity present.
6. Handle absences: Leadership: evaluate replacement and reassignment; escalate persistent deficits to senior management.
7. Evaluate and replan: Nursing, HR, and senior management: analyze metrics and return to staffing levels when needed.

Patient care hour deficit across four shifts

Example of four shifts in one unit in the same week: need/availability A=60/60, B=60/48, C=72/60, and D=48/48 hours. Values calculated as the maximum of need minus availability and zero. This is not a normative staffing calculation or a recommended distribution.

• Shift A: 0 patient care hours in deficit per shift
• Shift B: 12 patient care hours in deficit per shift
• Shift C: 12 patient care hours in deficit per shift
• Shift D: 0 patient care hours in deficit per shift

Example to illustrate the method.

Frequently asked questions

Should Cofen Resolution 543/2017 still be the basis for the schedule?

It was revoked by Resolution 743/2024. The official portal refers to Normative Opinion 1/2024 for planning parameters. Local documentation must identify the reference that applies to the service and be revised whenever the regulations are updated. [1][2]

Does dividing the number of patients equally balance the workload?

A headcount alone does not describe care needs, the concentration of activities, or the required competencies. Daily assignment should combine this information and be revisited when the demand profile changes.

Can scheduling software solve a staff shortage?

It can make conflicts visible and support distribution and replacement. A persistent lack of capacity requires a decision about headcount, organization, and service capacity; the system does not create available staff.

Sources

1. Resolução Cofen nº 743, de 12 de março de 2024
2. Parecer Normativo nº 1/2024/Cofen
3. Workload Indicators of Staffing Need: User Manual, 2nd edition, 2023

This article proposes an operational method for tackling time spent searching for supplies and equipment and the imbalance between shortage, excess, and expiration. The formulas and pilot described are management proposals to be adapted with supply chain, nursing, pharmacy, and clinical engineering. They do not represent an assessment of any specific hospital and do not replace technical decisions about products and care.

Define availability before designing the dashboard

Separate physical stock, reserved quantity, blocked items, and the balance actually available. A product in quarantine or with compromised packaging should not appear as an option for patient care just because it is still among the hospital's assets. The useful view for the unit combines item code, presentation, usable quantity, storage location, and time of the last update.

Consumables and mobile equipment require different records. For a gauze pad, track lot, expiration, and unit of use; for a wheelchair, individual identification, location, and operational status. An infusion pump may have been located and still require cleaning, maintenance, or technical release. Avoid a single field called available that hides these conditions.

Make the item master reflect real work

Start with the items that cause interruptions or losses, using a record approved by the responsible areas. Capture an unambiguous description, purchasing unit, distribution unit, conversion between package sizes, and authorized locations. Boxes and individual units cannot share a balance without an explicit conversion. Similar descriptions also need review to avoid duplicate records and picking errors.

The Ebserh quality manual (Ebserh is the federal company that manages Brazil's university hospitals) offers an institutional reference for inventory control, expiration, and losses; its requirements should be interpreted in the context of that network, without automatically turning them into a universal obligation. [1] As an implementation proposal, test the item master with the people who receive goods and the people who pick them: if two people interpret the presentation differently, the item master is not ready yet.

Track events, including transfers and returns

Each movement must specify item, quantity, origin, destination, person responsible, and time. Add lot and expiration where applicable. A transfer to a clinical supply cabinet changes where the stock is located; it does not prove the supply was used. If the institution records that issue as consumption for administrative purposes, document the limitation before estimating consumption per procedure.

The WHO recommends governance, financial sustainability, and identification and data exchange standards when structuring traceability of medical products. [2] A local application of this principle is defining which system is the source of record for each event. Duplicating the same issue in a spreadsheet and in the ERP creates two competing balances; an integration must reconcile confirmations, failures, and repeated entries.

Manage expiration with an action and an owner

Adopt an operational queue of products nearing expiration, sorted by date and by the possibility of authorized use. The FEFO approach, first expired, first out, only works when physical picking and the selection in the system follow the same lot. An alert without a recipient only changes where the backlog piles up.

For each alert, record a decision: keep the planned consumption, evaluate a permitted transfer, handle a contractual return, or send the loss to the institutional workflow. Compare the quantity that will expire with plausible demand before that date. Do not transfer items just to remove the risk from a local report, and do not automatically replace every product removed due to expiration: the stock level may be overstated.

Calculate restocking without treating the average as a guarantee

An initial reorder point can combine average daily consumption multiplied by the replenishment lead time and a safety stock set according to variability and criticality. For the unit cabinet, use the internal delivery time. For hospital purchasing, consider the full lead time until the product is accepted and available, including receiving and any required checks.

Worked example: consumption of 12 units per day, internal delivery in two days, and a safety stock of 18 units produce an initial reorder point of 42 units. These numbers are not a minimum stock recommendation. Reassess the assumption against delays, peaks, and storage capacity. Stockout days can artificially lower observed consumption: unmet demand needs to be recorded separately.

Use barcodes with a consistent identity

Anvisa's current page on UDI states that SIUD, Brazil's unique device identification system, has been in operation since March 2026 and sets out obligations with phased implementation. UDI identifies devices and supports surveillance activities; it is not the same as an automatic record of location or consumption inside the hospital. [3] Applicability must be checked by device and obligation, without assuming that all supplies already follow the same standard.

In the practice proposed here, the scanner must recognize the product, presentation, and available production information, and link them to the institutional item master. Internal labels must preserve the link to the original identification. Test scans on the packaging staff actually use, and define how to record illegible labels, split packages, and downtime without a scanner. The technology loses its value if the exception ends up as a note that is never reconciled.

Read the chart as an investigation queue

The chart shows an example of 50 requests that were not fulfilled immediately in one week: 18 due to an incorrect location, 14 due to a balance discrepancy, 10 due to delayed internal restocking, and 8 due to a shortage in central supply. Each request was assigned one primary cause.

In this example, buying more would not directly address the first three categories. Before deciding, it would be necessary to verify whether the causes were classified with consistent criteria and whether there are secondary causes. Do not use the proportions as a national estimate. The purpose of the chart is to show how to separate a location failure, a recording failure, and insufficient supply in order to choose a verifiable intervention.

Measure fulfillment and losses with clear denominators

Define a request as one requested line item, with an agreed quantity and deadline. A partial delivery does not count as full fulfillment. Record both the time to first delivery and the time to completion, so that a small delivery does not hide the outstanding balance. Emergency purchases also need a stable reason and classification.

Analyze stockouts and expiration losses together with average inventory and time spent searching for supplies. Fewer losses may result from lower stock, but also from incomplete write-offs; fewer requests may indicate that staff have stopped bothering to report shortages. Stratify by item, unit, and shift before comparing teams. The metrics below are proposed operational definitions, with no universal targets or promise of savings.

Test a complete cycle before expanding

Select one unit and a small group of supplies, covering receiving, storage, requests, restocking, consumption, and returns. Establish a baseline with a physical count and observation across shifts. Agree on who corrects the item master, who handles discrepancies, and who approves changes to stock levels. Include the staff who search for the items in the review of storage locations.

During the pilot, simulate a loss of connectivity, a transfer between units, and the discovery of an expired lot. Then compare records with physical evidence and examine the additional effort for the team. Only expand once movements can be reliably reconstructed and decisions on exceptions are documented. The next useful step is to deliver a list of confirmed problems and their causes, before choosing new tools.

Implementation checklist

• Check a sample of codes and box-to-unit conversions with receiving and clinical staff.
• Physically label storage locations and verify that they match the item master.
• Separate available, reserved, and blocked quantities in the inventory report.
• Test a transfer and a return while preserving lot and expiration.
• Assign an owner and a recorded decision to each expiration alert.
• Review actual internal delivery time before adjusting the reorder point.
• Reconcile the pilot's physical count and record the cause, correction, and approver of each discrepancy.

Tracking indicators

Full fulfillment on time

Request lines fully delivered by the agreed deadline / total lines with a deadline that ended in the period × 100.. Partial deliveries and lines still incomplete after the deadline stay in the denominator and out of the numerator.

Stockouts of monitored items

Item-day pairs with zero available balance / total item-day pairs tracked in the period × 100.. Define the items, units, and time of the check in advance; a zero balance alone does not measure unmet demand.

Expiration loss by value

Value of write-offs due to expiration in the month / value of average daily inventory for the same scope in the month × 100.. Use the same cost basis and disclose absolute values; the metric is unstable when average inventory is small.

Search time

Median minutes between the start of the search and locating the item in observed searches; also report the 90th percentile and the number of observations.. Record searches where the item was not found and the observation method separately; do not exclude them silently.

Process map: from receiving to restocking

1. Receive and inspect: Receiving records item, presentation, lot, expiration, and discrepancies.
2. Store and assign location: The storeroom confirms the location and segregates blocked products.
3. Request and pick: The unit reports demand; supply chain selects an eligible lot and records the issue.
4. Confirm transfer: The receiving unit checks quantity, identity, and destination.
5. Consume or return: The team records the corresponding event, preserving origin and reason.
6. Reconcile and restock: The manager investigates discrepancies, handles expiration, and reviews restocking levels.

Why wasn't the request fulfilled immediately?

Example week, 50 request lines not fulfilled immediately; one primary cause per line, with no overlap. The values add up to 50 and do not represent prevalence or the result of an intervention.

• Incorrect location: 18 requests
• Balance discrepancy: 14 requests
• Delayed internal restocking: 10 requests
• Central supply shortage: 8 requests

Example to illustrate the method.

Frequently asked questions

Do barcodes solve supply shortages?

They help identify and record events when the item master, scanners, and routines work. On their own, they do not fix supplier delays, inadequate stock levels, or storage without assigned locations.

Does every unit need to be tracked individually?

It depends on the product, the risk, and the purpose. Lot and expiration may be enough for consumables; mobile equipment requires individual identification. The level of detail should be defined with the technical leads.

Should every item near expiration be transferred?

No. A transfer depends on real need, storage conditions, integrity, transport, and institutional rules. The decision must avoid simply shifting the loss to another unit.

Sources

1. Manual de Requisitos do Programa Ebserh de Gestão da Qualidade, versão 03, 2022
2. Policy paper on traceability of medical products, 2021
3. UDI: Sistema de Identificação Única de Dispositivos Médicos

This article connects three operational problems: unreliable bed visibility, discharges with open issues, and bottlenecks in diagnostic tests and internal transport. The proposal is to track the entire chain, using observable events and clear boundaries of responsibility. The workflow presented is a management suggestion for institutional adaptation. The clinical decision to discharge and the care requirements remain the responsibility of the appropriate professionals.

The NIR coordinates interfaces that cut across departments

The Brazilian Ministry of Health describes the Internal Regulation Unit (Núcleo Interno de Regulação, NIR), a hospital patient flow office, as the structure that monitors the patient's journey from admission to discharge, working with external regulation and hospital teams. The guidance includes communication between departments and agreements with diagnostic services. This reference covers the complete flow. [1]

Define monitoring and execution. The clinical team confirms discharge conditions and needs; the services involved are accountable for their respective deliverables; coordination tracks blockers and escalates to the appropriate authority. Without agreements between departments, the NIR receives demands it cannot resolve.

Separate the patient's status from the bed's status

Propose two related sets of information. For the patient's journey, track expected discharge, the recorded decision, open issues, and actual departure. For the physical resource, track occupancy, waiting for cleaning, cleaning in progress, blocked, available, and reserved. Forecasts change and reservations constrain assignments; preserve the history.

Each transition needs a definition and someone who confirms it. Actual departure is the moment the patient leaves the bed according to the institutional criterion, not the time someone closes a record. Available means the local requirements for use have been met. Where possible, record the time of the event and the time of the update separately. The difference reveals information lag.

Discharge barriers must surface before departure

In the multidisciplinary review, use the expected discharge as a planning hypothesis, with a date and reason for revision. Identify paperwork, patient education, continuity of care needs, coordination with the care network, and transport, where applicable. For each dependency, record the owner, the next action, and when it will be reassessed. Distinguish what can already be prepared from what depends on clinical confirmation; getting ahead administratively does not justify rushing the clinical decision.

If a diagnostic test is holding up the decision, track its actual sequence: order, required preparation, collection or transport, performance, availability of the result, and review by the responsible team. This breakdown distinguishes equipment unavailability, preparation, the diagnostic queue, and a result not yet reviewed. The classification should describe the dependency observed, without turning every delay into an individual failure.

Cleaning completed and bed released are separate events

Anvisa's manual on surface cleaning and disinfection distinguishes cleaning processes and places terminal cleaning in the context of room vacancy and scheduled routines. It is a technical reference, published in 2010 and still made available by the agency, that must be aligned with local procedures and applicable health regulations. A management turnaround target does not replace these technical conditions. [2]

Record the cleaning request, start, and finish; then confirm the institutional release requirements and any blockers. The Coren-BA opinion (Coren-BA is the Bahia State Nursing Council), revised in 2024 and published in 2025, sets out, in its regional context, the interface between the removal of supplies by nursing and terminal cleaning by the trained environmental services team. It is a regional reference for discussing responsibilities, not a national rule. [4]

Internal transport needs readiness and a confirmed destination

For tests or internal transfers, an operational queue can show the request time, the priority set by the appropriate team, origin, destination, conditions required for the move, and request status. Before dispatch, confirm that both origin and destination are ready according to the service protocol. Recording only that the request was fulfilled hides returns, cancellations, and waits at a unit that cannot yet receive the patient.

Distinguish internal transport from transport out of the hospital, which may involve other providers and access conditions. Assign one identifier per request to avoid duplication and keep the reason for any cancellation. When the clinical situation changes, the appropriate team reassesses the priority and whether the patient can be moved. Tracking locates dependencies without replacing a safety assessment.

Turnover, occupancy, and step times answer different questions

The São Paulo State Department of Health portal brings together metrics for operational beds, occupancy, length of stay, turnover interval, and bed turnover. Its definition of operational beds is based on the daily census, including occupied and vacant beds in operation and excluding blocked beds. This attention to the denominator is essential when capacity changes during the month. [3]

To track turnover, state which discharges are included in the numerator, the period, and how the average number of operational beds is calculated. To find where the process waits, use event timestamps: departure to cleaning request, request to start, execution, and release. Look at the median and 90th percentile of each interval, plus the proportion of records missing a timestamp. The aggregate turnover interval is not the same as the cleaning time for each bed. Do not compare different specialties without considering the care profile and available resources.

Worked example: break it down before choosing an intervention

The chart represents a single example path after the discharge decision: 70 minutes until actual departure, 20 until cleaning is requested, 25 until it starts, 35 of cleaning, and 10 until release. The intervals are sequential and total 160 minutes. The first three steps add up to 115 minutes, but they cannot automatically be classified as avoidable time: the conditions of each transition need to be checked.

The example raises different questions. Did the period before departure depend on patient education, paperwork, or transport? Was the cleaning request late because the system did not receive the event? Was capacity short at peak time? Analysis comes before the solution. These numbers illustrate the method and are not an average, a target, or a benchmark. For a real series, do not add up the medians of each step as if they were necessarily the median of the total path.

Test the reliability of status data before automating follow-ups

Start by reconciling the dashboard with observation on the unit at different times of day. Classify discrepancies and identify who can correct each piece of information. Then test a workflow on one unit: confirmed event, owner notified, completion recorded, and exception reviewed. Once the statuses are understandable and the timestamps reliable, evaluate notifications for agreed delays. Alerts based on wrong data only speed up misdirected follow-ups.

Also examine cleaning rework, reopened blocks, transport cancellations, and reports of patients leaving without adequate discharge education. These signals balance speed and quality. The next step is to bring together the NIR, clinical care, environmental services, diagnostic services, and transport to choose one verifiable dependency and one small change. A lack of capacity or continuity in the care network requires a management decision and external coordination, not just rearranging the colors on the dashboard.

Implementation checklist

• Distinguish expected discharge, discharge decision, and actual departure.
• Define criteria for availability, reservation, and blocking.
• Assign an owner and a next action to each open issue.
• Record cleaning request, start, finish, and bed release.
• Separate internal transport, diagnostic tests, and transport out of the hospital.
• Check the dashboard against the observed situation on the unit.
• Set the period, numerators, denominators, and exclusions.
• Review waits, rework, cancellations, and continuity of care.

Tracking indicators

Bed turnover in the period

Eligible hospital discharges ÷ average daily operational beds in the period.. State how discharges, deaths, and external transfers are handled; exclude internal transfers. Daily average = sum of operational bed-days ÷ days. Higher turnover does not prove appropriate discharge.

Operational occupancy (%)

100 × patient-days ÷ operational bed-days in the period. [3]. Keep the census method consistent and make blocked beds explicit.

Time from departure to release

Release time minus actual departure, per episode.. Report the median, 90th percentile, valid episodes, and exclusions. Not the same as cleaning time.

Wait for cleaning to start

Cleaning start minus request, per episode.. Report the median and 90th percentile by shift; investigate readiness and capacity.

Wait for internal transport

Start of transport minus valid request.. Report the median and 90th percentile by priority and destination; identify cancellations.

Status agreement (%)

100 × beds with confirmed status ÷ beds audited.. Fixed criteria and varied times. Proposed data quality metric.

Process map: from expected discharge to the next admission

1. Plan: The team estimates departure and identifies dependencies.
2. Confirm and prepare: Appropriate professionals record the decision; owners track deliverables.
3. Confirm departure: The unit records that the room is physically vacated and prepares it.
4. Clean: The team records request, start, finish, and obstacles.
5. Check and release: The owner verifies requirements; obstacles keep a justified block in place.
6. Assign and transport: Patient flow and clinical staff confirm suitability and readiness.
7. Confirm occupancy: The unit records arrival; coordination reviews exceptions.

One path from discharge decision to bed release

Sequential path with no overlap: 160 minutes to release. Does not include the next admission and does not set a target.

• Discharge decision to departure: 70 minutes
• Departure to cleaning request: 20 minutes
• Request to cleaning start: 25 minutes
• Cleaning: 35 minutes
• End of cleaning to release: 10 minutes

Example to illustrate the method.

Frequently asked questions

Does a recorded discharge mean the bed is available?

No. Actual departure, cleaning, and institutional release criteria are separate steps.

Is there an ideal release time?

No universal deadline is set here. Local goals must preserve technical requirements and safety.

Does higher turnover prove improvement?

No. Examine the care profile, operational capacity, appropriateness of discharge, and rework.

What if a diagnostic test is holding up departure?

Pinpoint the dependency, from preparation to review of the result, without preempting the professional decision.

Sources

1. Política Nacional de Atenção Hospitalar: regulação de leitos intra-hospitalares
2. Segurança do paciente em serviços de saúde: limpeza e desinfecção de superfícies
3. Indicadores e Relatórios de Unidades: conceito e fórmula de cálculo
4. Atualização do Parecer Técnico Coren-BA nº 007/2018

This guide uses as an example the launch of an occupational health line inside a hospital: visits with an occupational medicine physician, audiometry, lab and imaging tests, and organizational psychology support. The same reasoning applies to other specialties. The recommendations are operational and do not replace reading the current regulations or guidance from the responsible clinical leads.

Start with an inventory, not a date

A midsize hospital or clinic almost always has much of what a new service needs: exam rooms, a front desk, a lab with a draw station, X-ray, and scheduling and billing systems. The first task is to list, requirement by requirement, what the new service demands and compare it with what exists. The comparison has to be concrete: it is not enough to say there is a lab; you need to know whether the draw station can handle more patients at the time of day employers want to send their workers.

The World Health Organization's service availability and readiness assessment, known as SARA, treats readiness as the availability of the components needed to deliver a service and measures service-specific readiness through tracer items such as trained staff, guidelines, equipment, diagnostics, and medicines. [4] The tool was designed for surveys of health systems, but the logic carries over to a new service: the question is not whether the hospital is good, but whether every component this service requires is available and working.

Check the formal requirements before announcing

For occupational health in Brazil, the starting point is NR-7, the regulatory standard for the Occupational Health Medical Control Program (PCMSO), which is tied to each employer's risk assessment under its Risk Management Program (PGR). The standard requires pre-employment, periodic, return-to-work, change-of-exposure, and exit exams, each made up of a clinical exam and supplementary tests, and an Occupational Health Certificate (ASO) issued for every clinical exam. [1] The employer is the one that contracts the PCMSO; the healthcare service must be able to perform the tests the program specifies.

Annex II of the same standard sets specific requirements for audiometry: an audiometric booth or acoustically treated room that meets ISO 8253-1, an annual acoustic check of the audiometer, periodic calibration, testing performed by a physician or audiologist, and at least 14 hours of auditory rest before the test. [1] Details like these determine the equipment, the room, the professional, and even the instructions given to the worker at scheduling.

Two other checks apply to any specialty. If the launch involves construction, expansion, or renovation, Anvisa's RDC 50/2002 requires the local health surveillance authority to review the physical design before work begins. [2] And Brazil's National Registry of Health Facilities (CNES) records the physical resources, workforce, and services of each facility, with mandatory updates whenever that information changes. [3] Regulations get revised; check the current text and the municipality's requirements before locking in the timeline.

Classify each gap by the type of fix it needs

Once the inventory is done, each requirement gets one of the four classifications below. The classification matters because each type of gap has a different timeline, cost, and owner. Adjusting a schedule takes days; hiring a professional or buying equipment can take months.

• Exists and meets the need: requirement available, with evidence such as a contract, certificate, or staffing schedule.
• Exists, needs adjustment: resource available, but with hours, capacity, or workflow that do not fit the new service.
• Does not exist, hire or partner: a professional or service that makes sense to bring in from outside.
• Does not exist, buy or build: equipment, a room, or a tool that has to be purchased or set up.

Map the journey of the people you will serve

An occupational health service has two clients: the employer, which schedules and pays, and the worker, who shows up. Map the worker's journey from scheduling to delivery of results: how the employer sends the list of tests specified in the PCMSO, where the worker checks in, in what order they go through the draw station, X-ray, audiometry, and the physician visit, and who pulls everything together so the physician can issue the certificate. A poorly planned sequence sends the worker back to the front desk twice or has them wait for audiometry after hours in a noisy environment.

The journey also shows where the new service competes for capacity with the existing one. If all the workers arrive at seven in the morning, the draw station and X-ray that serve outpatient and inpatient care get an extra peak. Holding time windows for the new line is a scheduling decision that has to be made before opening, together with whoever runs those departments.

Lock in partners, costs, and pricing assumptions with finance

Partnerships, such as organizational psychology or audiology, need scope, a responsible clinical lead, a scheduling method, and a delivery deadline defined in writing. On the financial side, build the cost of each package from its components: visit time, draw supplies, lab and imaging tests at internal cost, partner fees, certificate issuance, and a share of the shared overhead.

The price quoted to the employer should be based on these assumptions, recorded in a spreadsheet that finance signs off on, with estimated monthly volume, test mix, and payment terms. When actual volume differs from the forecast, the assumptions show what changed. Without them, the margin only comes to light months later, at the financial close.

Treat results turnaround as a commitment to the client

For the employer, the service is done when the certificate and the results arrive. A late pre-employment exam delays a hire. That is why the delivery time should be calculated from the real workflow: lab turnaround, X-ray report time, and the physician's availability to review everything and issue the certificate. Only then does it go into the commercial proposal, with a clear rule for tests that need to be repeated or supplemented.

Prepare the team and set go-live criteria

The front desk, draw station, imaging, and billing will serve a new population, with identification, authorization, and billing rules that differ from those for health plans or the SUS. A short guide for each role, a dry run with mock workers, and clarity on who handles exceptions cut down on questions in the first week.

Go-live criteria are verifiable conditions that must be met before the first client is seen: equipment checked and certified, professionals hired and registered, workflow tested in a dry run, delivery time calculated, price signed off by finance, registry updated. The opening date becomes the date the last criterion is met, not the other way around.

Practical example: 20 go-live requirements

Picture a hospital that lists 20 requirements for its occupational health line. Eleven already exist and meet the need, including exam rooms, the front desk, an in-house lab, X-ray, and the scheduling and billing systems. Five exist but need adjustment: a time window on the X-ray schedule, extra capacity at the draw station during the morning peak, a results delivery format for employers, a front desk workflow for workers, and an update to the service's registry entry. Two require hiring or a partnership: an occupational medicine physician and an audiologist. Two need to be bought or set up: the booth with a checked audiometer and the process for issuing and storing certificates.

The chart shows this breakdown. Sixteen of the twenty requirements rely on what the hospital already has, and that is what makes the launch feasible. But the four missing items set the date: until audiometry is set up and the occupational medicine physician is hired, no promised date is safe.

Monitor the first few weeks closely

The first few weeks show whether the assumptions hold. Track weekly the actual delivery time against what was promised, how long workers spend on site, the open items that prevented a certificate from being issued, and volume against the pricing assumption. A short weekly meeting with the front desk, draw station, imaging, the responsible physician, and finance makes it possible to fix the workflow before problems turn into complaints from the employer. Once things stabilize, monitoring can move to a monthly cadence.

Implementation checklist

• List the new service's requirements and compare each one with what already exists.
• Classify each gap as an adjustment, a hire or partnership, or a purchase.
• Check the applicable regulations and local requirements in their current text.
• Map the journey of the person served, from scheduling to delivery of results.
• Validate cost per package, volume, and pricing assumptions with finance.
• Calculate delivery time from the real workflow before putting it in the proposal.
• Run a dry run with the team and decide who handles exceptions.
• Open only when every go-live criterion has supporting evidence.

Tracking indicators

Go-live readiness

100 × go-live requirements met with evidence / total requirements defined for the service.. A requirement without evidence, such as a certificate, contract, or staffing schedule, counts as open. High readiness does not offset a single critical requirement that is still open.

Results delivery time

Median and 90th percentile, in business days, from the worker's visit to delivery of the certificate and results to the employer, compared with the contracted turnaround.. The 90th percentile reveals the cases that lead to complaints. Track tests that had to be repeated or supplemented separately.

Time on site

Median, in minutes, from the worker's arrival at the front desk to discharge after the last test of the day.. Increases point to conflict with the existing service's workflow or a poor test sequence. Analyze by time slot.

Volume against the pricing assumption

100 × visits completed in the month / monthly volume used in the pricing assumption signed off by finance.. Persistent gaps call for a review of cost per package and price. This is not a sales target; it is a check on the assumption.

Process map: from opportunity to opening a new service

1. Inventory what exists: Management lists the service's requirements and compares them with available rooms, staff, lab, and imaging.
2. Classify gaps: Each requirement gets a fix type, an owner, and an estimated timeline.
3. Map the journey: Workflow from scheduling to delivery of results, with time windows held in shared departments.
4. Lock in partners and pricing: Contracts with scope and deadlines; cost per package and assumptions signed off by finance.
5. Prepare the team: Guide for each role, a dry run of the visit, and an owner for exceptions.
6. Verify go-live criteria: Open only with evidence for each criterion; the date follows from the last item met.
7. Monitor the first weeks: Weekly meeting on delivery time, time on site, open items, and volume against the assumption.

Status of 20 go-live requirements for an occupational health line

Example: 11 requirements already exist and meet the need; 5 exist but need adjustment; 2 require hiring or a partnership; 2 need to be bought or set up. Total: 20 requirements.

• Exists and meets the need: 11 go-live requirements
• Exists, needs adjustment: 5 go-live requirements
• Does not exist, hire or partner: 2 go-live requirements
• Does not exist, buy or build: 2 go-live requirements

Example to illustrate the method.

Frequently asked questions

How long does it take to launch an occupational health service?

It depends on the gaps. If almost everything already exists, the timeline is driven by scheduling adjustments and team preparation; if professionals or equipment are missing, it is driven by hiring and procurement. That is why the inventory comes before the date.

Does the hospital need to perform every test in-house?

Not necessarily. Some tests can be done by partners, as long as scope, responsible clinical lead, scheduling, and delivery time are defined and the regulatory requirements, such as those for audiometry, are met. [1]

Who decides which tests each worker gets?

The tests follow from the employer's PCMSO, built from the risks classified in the PGR, under the responsibility of the physician in charge of the program. The healthcare service performs the tests and delivers the results within the agreed time. [1]

Sources

1. Norma Regulamentadora nº 7 (NR-7): Programa de Controle Médico de Saúde Ocupacional, PCMSO
2. Resolução RDC nº 50, de 21 de fevereiro de 2002: regulamento técnico para planejamento, programação, elaboração e avaliação de projetos físicos de estabelecimentos assistenciais de saúde
3. Portaria de Consolidação nº 1, de 28 de setembro de 2017, arts. 358 a 373 (origem: Portaria GM/MS 1.646/2015, Cadastro Nacional de Estabelecimentos de Saúde)
4. Service availability and readiness assessment (SARA)

Managing metrics and costs in healthcare means answering three questions: what is being measured, which records support the calculation, and what decision will be made. This guide proposes a minimum set of definitions, controls, and responsibilities to address unreliable manual metrics and opaque costs. The operational recommendations need to be adapted to each service.

Start with the decision the metric should support

Choose a well-defined management question: why did the cost per patient day change in this unit? Define who can act on the result, in which meeting, and with what supporting evidence. Without that link, the dashboard piles up measures that no one can turn into a verifiable action. A cost reduction must also be examined alongside the service's capacity and quality.

The Brazilian Ministry of Health presents the National Cost Management Program (PNGC) as support for systematic cost accounting, the structuring of cost centers, and the use of ApuraSUS, its cost accounting system. The program covers monthly costs and average unit costs of each department's outputs. This provides an institutional reference for organizing the discussion, without automatically making facilities with different care profiles comparable. [1]

Write a metrics dictionary that can actually be executed

Each definition sheet should state the name, purpose, numerator, denominator, unit of measure, included population, exclusions, period, source, owner, and version. Add the rounding rule, handling of missing data, closing date, and review criteria. These fields are a practical governance proposal. The WHO includes standardized lists and denominator challenges among the foundations of health facility data analysis. [2]

For cost per patient day, specify which costs go into the numerator and how output is counted each day. Do not substitute the number of admissions for patient days. Define how transfers and records corrected after the close are handled. An executable definition sheet allows two people to reproduce the result without turning to the author of the original spreadsheet. If the rule changes, record the effective date and assess the break in the historical series.

• Definition owner: validates meaning, inclusions, and interpretation.
• Data owner: confirms origin, coverage, and timeliness.
• Calculation owner: runs the formula and keeps the version used.

Validate coverage and consistency before interpreting performance

The WHO organizes data review into dimensions such as completeness, internal consistency, and external comparisons. [3] In local practice, this can become a simple sequence: check expected files, required fields, duplicates, and consistency between totals and subtotals. An unusual value should trigger an investigation. It should not be quietly corrected to bring the result closer to the target or to the previous month.

Missing information is not the same as zero. If a unit has not submitted its output, the unit cost should remain pending, with the reason visible. If the denominator is zero, do not present an artificially adjusted division. Separate provisional data from closed data and keep a log of corrections. Before the meeting, report which sources are complete and which limitations could change the conclusion about the period.

Connect costs and output with explicit criteria

Propose a matrix linking cost centers, supply items, staffing, contracts, and output. Costs identifiable with a department can be assigned directly; shared expenses require a documented criterion. Floor area, processing volume, or another operational basis may be candidates, depending on the nature of the expense. The choice must make sense for that resource and be approved by the controller's office, with an analysis of its limitations.

Purchase, consumption, payment, and allocated cost are not equivalent events. A purchased supply may remain in inventory; a payment may relate to a different accounting period. Preserve the link between source documents, movements, and the amounts allocated to the month. The institutional method should also spell out the treatment of depreciation, outsourcing, and internal transfers. This article does not replace the accounting policy or define an official costing methodology.

Worked example: unit cost can rise without an increase in the total

Consider a unit with an allocated monthly cost of R$ 300,000 over three months. Output is 600, 500, and 625 patient days, respectively. The average cost is R$ 500, R$ 600, and R$ 480 per patient day. The chart shows these calculations, keeping the same cost scope.

In the second month, output falls and the unit cost rises, even though the total cost stays the same. This justifies investigating available capacity, cost structure, and changes in the care profile. It does not demonstrate waste or justify cutting staff. Before comparing units, check the set of services included and differences in complexity. Also, do not confuse average accounting cost with savings that could actually be achieved by reducing output by one unit.

Turn the cost close into a process backed by evidence

The process map below starts with the sources and ends with a tracked decision. For each step, set an observable output: batch received, inconsistency report, calculation worksheet, approval, and recorded action. Archive the version of the formula and the time of extraction. This makes it possible to explain why a published value changed, without turning the history into a series of overwritten spreadsheets.

A cost review should reconcile the source amounts with the final allocations. Differences need an owner, a justification, and a defined treatment. On the executive dashboard, show the result together with the update date and the validation status. Keep detailed financial data in an authorized environment, using aggregates in broader discussions. No example in this guide requires a patient's name, diagnosis, or any other identifiable data.

Test the routine before expanding the dashboard

Start with one unit and one closed period. Redo the calculation with the source owners and compare the manual calculation with the proposed routine. The acceptance criterion can be reproducing the result from the same records, with every difference explained. This implementation choice is an operational recommendation, not a mandatory deadline or a promise of benefit.

At the next meeting, check whether the decision produced an owner, a deadline, and evidence of completion. Record the questions that require new data collection instead of immediately multiplying metrics. A useful dashboard lets people challenge the calculation with documents and revisit the interpretation. The next step is to stabilize this closing routine and only then bring in other units, preserving the differences in scope that make each comparison valid.

Implementation checklist

• Select one decision, one unit, and one reporting period.
• Approve definition sheets with formula, source, exclusions, owners, and version.
• Distinguish a zero value, missing data, and an invalid denominator.
• Reconcile source amounts, allocations, and output for the accounting period.
• Document the criteria for allocating shared costs.
• Flag provisional results and keep a record of revisions.
• Link each decision to an owner, a deadline, and evidence of follow-up.

Tracking indicators

Average cost per patient day

Costs allocated to the unit in the month, within the approved scope, divided by the daily sum of inpatients in the same unit and month; unit: R$/patient day.. A zero denominator produces a result that cannot be calculated. Compare only equivalent scopes; this does not estimate marginal cost or the savings from a cut.

Completeness of required fields

100 × required fields validly filled in / required fields expected in the batch records; exclude non-applicable fields according to a documented rule.. High completeness does not prove accuracy. Also show the number of records received versus expected.

Source timeliness

100 × deliveries received by the agreed deadline / deliveries expected in the period.. The deadline is a local agreement; a single late source can jeopardize the close even when the others arrive on time.

Reconciliation difference

Total source costs within scope minus total allocated after allocations, using the same accounting period and adjustment rules; unit: R$.. Every difference requires an explanation. A zero difference can also hide inappropriate classifications or allocations.

Process map: from source to management decision

1. Receive sources: Source departments deliver supply, staffing, contract, and output data, with accounting period and version.
2. Validate batches: Data owners check coverage, duplicates, and consistency; open issues go back to the source.
3. Calculate and reconcile: The controller's office applies approved criteria and documents the link between costs and output.
4. Approve the close: Management and definition owners validate the calculation worksheet or send back inconsistencies.
5. Decide and follow up: The owner records the action, deadline, and evidence; findings feed into the next close.

Cost per patient day with constant total cost

Example: R$ 300,000 in total cost per month. Denominators: month A, 600 patient days; month B, 500; month C, 625. Values calculated by division.

• Month A: 500 R$/patient day
• Month B: 600 R$/patient day
• Month C: 480 R$/patient day

Example to illustrate the method.

Frequently asked questions

Can we start with spreadsheets?

Yes, as long as there is an identified source, an approved calculation rule, version control, and review. Automation should preserve these controls and reduce repetitive tasks.

Does a lower cost per patient day mean better performance?

Not on its own. Changes in output, care profile, and cost scope can alter the result. Interpretation must consider quality, capacity, and comparability.

How should a metric with a missing denominator be handled?

Keep the result pending and identify the missing source. Do not turn missing data into zero or use an estimated denominator without an explicit rule and flag.

Sources

1. Programa Nacional de Gestão de Custos
2. Toolkit for analysis and use of routine health facility data: general principles
3. Data quality assurance: module 2: discrete desk review of data quality

This guide covers sources of truth, integration contracts, incidents, and traceability without presenting identifiable clinical data. The scenarios are generic and the chart uses a worked example. Brazil's Ministry of Health defines the RNDS (National Health Data Network) as the national infrastructure for standardized sharing of health information. This initiative is a reference point for interoperability, but internal workflows still require local design and clearly assigned responsibilities. [1]

Define sources of truth by information domain

A source of truth is the system authorized to maintain a given set of information, with an owner and a correction rule. An organization may choose an enterprise master registry for identifiers, the pharmacy system for inventory movements, and a dedicated log for incidents. This allocation is a proposed operational architecture. It should be confirmed with the process owners, without assuming that any one application has authority over all data.

For each domain, document who creates, who consults, who can correct, and how the change reaches downstream consumers. When two systems disagree, the rule should indicate where to investigate and who decides. Do not resolve conflicts by always choosing the most recently received record: an older message may have been held up. Keep the version and the time the event occurred separate from the time it was received.

Write an integration contract that includes meaning

Describe the event that triggers the exchange, required fields, identifiers, units, terminologies, possible states, and the expected acknowledgment. Add version, destination, owner, and handling of mismatches. A technically well-formed message can still carry a code with a different meaning than expected. That is why acceptance must check content and effect on the process, in addition to the format received.

To avoid duplicates, one approach is to assign a stable key to each event and make repeated processing unable to generate a second, improper transaction. Also define how out-of-order messages are handled and set a limited number of retries. Persistent failures should go to a review queue with an assigned owner. These decisions are design recommendations; they depend on system capabilities and do not describe a universal vendor feature.

Follow the medication workflow to the correct step

In the medication workflow, distinguish between a recorded prescription, an amended version, dispensing, and documented administration. One state does not prove the next. The integration should make it possible to detect when a revision has not propagated and route the discrepancy to the authorized team. Clinical decisions remain with the responsible professionals; the integration mechanism should not reinterpret or complete a prescription to work around a technical failure.

Before go-live, rehearse a change followed by a cancellation, a system outage, and a resend of the same event. Use synthetic records in a test environment. Define how the team works during downtime and how records will be reconciled afterward. Restoring the connection is not enough: you need to verify that the correct sequence was recovered and whether there are pending items that could affect care.

Separate information provenance from access history

The HL7 FHIR R4 specification describes Provenance as a way to record the context in which a resource was created or revised and the agents involved. This resource helps investigate where a piece of information came from. [3] AuditEvent, in turn, is primarily intended to record events for security auditing, including system usage activity. [4] These are distinct conceptual references, not automatic requirements for every internal integration.

For an operational trail, consider storing the event identifier, sending system, version, timestamp, result, and correlation identifier. Preserve failures and corrections under access control, and avoid deleting history to make the dashboard look clean. Metadata can also identify people or allow indirect linkage. Therefore, restricting clinical content does not eliminate the need for protection, a retention policy, and an analysis of what actually needs to be recorded.

Link incident reporting to investigation and feedback

An integration failure may be purely a technical issue or part of a patient safety incident. Classification requires a process review, with participation from the responsible departments and the Patient Safety Committee (Núcleo de Segurança do Paciente) when appropriate. The initial form can capture minimal context and allow details to be added later, reducing effort without compromising the investigation. Link the technical ticket to the safety report through a controlled reference, avoiding unnecessary copies.

Anvisa, Brazil's health regulatory agency, maintains reporting guidance, tutorials, and update notices for the Health Care module of Notivisa, its adverse event reporting system. [5] The internal workflow should support assessment and any applicable reporting while preserving institutional responsibilities. Do not assume that opening a ticket or integrating a tool already submits the report to the agency. Close the loop with a corrective action, an owner, verification, and feedback to the team that reported the problem.

Distinguish communication sent, delivered, and understood

For instructions given to patients and families, separate drafting, review, sending, delivery, and confirmation of understanding. Each state answers a different question. A channel can confirm delivery without showing that the instructions were current or that the recipient was authorized. Define which version is valid and who may revise it. Maintain institutional channels and an alternative when there are barriers to access or comprehension.

Brazil's data protection law (LGPD) classifies health-related information as sensitive personal data and sets out principles such as purpose limitation, necessity, and security. [2] In the proposed design, use only data relevant to the purpose, with role-based access and an institutional assessment of the applicable legal basis. Replacing names with codes does not guarantee anonymization. For broad dashboards, prefer aggregates with a reidentification risk assessment; for testing and demos, use synthetic data.

Measure the integration outcome and test recovery

In the example in the chart, a batch has one thousand events: 950 processed successfully on the first attempt, 30 processed after a retry, and 20 pending at cutoff. The categories are mutually exclusive. All events may have been received, but only 980 had confirmed processing. The completion rate is 98%, with 2% still pending. These values are not a recommended target.

Beyond the percentage, track the age of the oldest pending item and the propagation time. Completed events alone do not reveal how delayed the stuck ones are. Start with a well-scoped workflow, with an identified operational owner and technical support. Acceptance should include failure, recovery, reconciliation, and a demonstration of the final state at the destination. Only then expand the scope, keeping stop criteria in place when there is risk to the process.

Implementation checklist

• Map each domain, its authorized source, and who is responsible for corrections.
• Document fields, semantics, versions, and states for each integration.
• Distinguish technical receipt from confirmed processing at the destination.
• Rehearse duplication, delay, cancellation, and downtime using synthetic data.
• Define downtime procedures and reconcile pending items after recovery.
• Link technical issues and patient safety incidents with controlled access.
• Verify recipient, version, and status of institutional instructions.
• Review purpose, access, retention, and reidentification risk of records.

Tracking indicators

Batch processing completion

100 × unique events with confirmed business processing by cutoff / eligible unique events issued in the batch.. Define eligibility, cutoff time, and confirmation. A transport-level response does not prove the process was completed.

P95 propagation time

95th percentile of the time between issuance and confirmed processing, for completed events in the defined cohort; unit: seconds or minutes.. Report the percentile method, clock synchronization, and pending events. A cohort of completed events only can hide persistent failures.

Pending items at cutoff

100 × eligible unique events without final confirmation at cutoff / eligible unique events in the batch.. Show absolute count, age, and criticality. Define separate rules for cancellations and final rejections.

Verified corrective actions

100 × actions due in the period with verified effectiveness / actions due in the period, according to criteria defined by the process owner.. Closing a ticket does not demonstrate effectiveness. This metric does not replace safety analysis or reporting obligations.

Process map: from event to confirmation and reconciliation

1. Generate event at the source: Domain owner confirms the event's identifier, version, and meaning.
2. Validate and route: Integration layer checks the contract and logs correlation, without exposing clinical content on dashboards.
3. Process at the destination: Consuming system checks for repeats, ordering, and business rules; returns a confirmation or failure reason.
4. Handle exceptions: Support and the operational owner review pending items; clinical assessment triggers the safety workflow when appropriate.
5. Reconcile and give feedback: Departments confirm the final state; gaps go back for review and corrective actions are verified.

Final status of an integration batch

Batch of 1,000 unique events at cutoff. Exclusive categories: 950 completed on the first attempt, 30 after a retry, and 20 pending. Total confirmed: 980 events, or 98%.

• Completed on first attempt: 950 unique events
• Completed after retry: 30 unique events
• Pending at cutoff: 20 unique events

Example to illustrate the method.

Frequently asked questions

Does using FHIR resolve every discrepancy between systems?

No. A standard organizes representation and exchange, but you still need to define meaning, versions, identifiers, responsibilities, and confirmation of processing in the local workflow.

Does a technical ticket replace incident reporting?

No. The ticket and the safety report serve different purposes. The institution must assess the event and follow the applicable reporting guidance and responsibilities. [5]

Can we copy real data to test an integration?

This guide recommends synthetic data. Processing real data requires an institutional assessment of purpose, legal basis, necessity, and safeguards, especially for health data. [2]

Sources

1. Rede Nacional de Dados em Saúde
2. Lei nº 13.709/2018, Lei Geral de Proteção de Dados Pessoais
3. FHIR R4: Resource Provenance
4. FHIR R4: Resource AuditEvent
5. Como notificar incidentes/eventos adversos relacionados à assistência à saúde

This article covers three connected problems: incomplete shift handoffs, interruptions and scattered requests, and open care tasks without follow-up. The goal is to organize nursing work without turning the conversation into a full read-through of the chart. The workflow and metric proposals below are operational adaptations for institutional review, not clinical orders or a promise to reduce adverse events.

What needs to carry over at shift change

AHRQ describes the handoff as a structured transfer of information, authority, and responsibility. Its material emphasizes recent changes, uncertainties, the plan of care, contingencies, and the opportunity to ask questions, with acknowledgment by the receiver. This guidance helps evaluate the handoff by the understanding achieved, not just by whether a form exists. [1]

In practice, consider three questions per patient: what has changed since the last assessment, what needs attention in the next period, and what is still open? The context should make the priority of each open item clear. A long task list without that context can make different situations look equally urgent. Clinical content and decisions remain subject to professional judgment and institutional protocols.

How to use SBAR without creating a bureaucratic read-through

SBAR organizes communication into situation, background, assessment, and recommendation or request. The official AHRQ tool includes stating explicitly what is needed and when, with a read-back of the response to confirm accuracy. It is a reference for structuring the conversation, not a guarantee that every risk will be identified. [2]

For the shift, select the information relevant to each situation and add a list of open actions. An example: an assessment has already been requested, the department has not yet confirmed it, and the receiving nurse needs to check for a response at the agreed time. The example shows how communication is organized, without defining care for any patient. At the close, the receiver summarizes priorities and clarifies discrepancies before taking over follow-up.

An open item needs an origin, an owner, and a closure

A useful operational tracking sheet contains a link to the encounter, the expected action, the source of the request, the accountable professional or team, the agreed deadline, and the current status. Suggested statuses: open, accepted, in progress, blocked, completed, or canceled with justification. A blocked status must state which dependency is preventing progress and who should be contacted. Assignments must respect professional scope of practice; agreeing to follow up on a dependency does not transfer another discipline's legal responsibilities.

Define what proves completion for each type of task. Calling a department may close the contact action, but not the pending assessment. COFEN Resolution No. 736/2024 (from Brazil's Federal Nursing Council) requires formal documentation of the Nursing Process in the paper or electronic health record and distinguishes entries by registered nurses from those by other team members. A tracking list should support that documentation, not replace it. [3]

Scattered requests need to enter a recognized workflow

Before choosing a tool, observe where requests come from: phone, hallway, system, and messages. Over a sample of shifts, record only the categories and times needed for the analysis. Check how many requests require a second search because no one knows who received them. This assessment distinguishes a missing channel, insufficient capacity, and interruptions caused by a real change in the patient's condition.

As a local proposal, set up an institutional intake point for non-urgent requests and a point person during handoff, as staffing allows. Emergencies follow the escalation process defined by the service and must interrupt routine when necessary. Afterward, the team reconciles what was left open. Protecting the handoff means reducing avoidable fragmentation, without creating a barrier to critical alerts or requiring anyone to wait in an administrative queue to ask for help.

From preparation to the check at the start of the new shift

The proposed map starts with updating documentation and reconciling open actions. Then come a prioritized summary, questions, confirmation, and an initial review by the incoming team. Each step must produce an observable output: a reviewed list, questions answered, or an owner assigned. If the conversation and the record disagree, flag the item for clarification, keeping the previous information until the correction is justified.

The handoff should also accommodate absences, unit transfers, and changes of owner mid-shift. In these cases, record who took over and when, using the institutional mechanism. If the system is down, use the approved downtime procedure and reconcile notes when it comes back. Avoid multiple independent copies that keep circulating after an update, especially when they include clinical information.

Metrics that reveal breaks in continuity

Measure the share of audited handoffs with the local required fields completed, open items with an owner and a deadline, and overdue items still open. Fix the time of data collection: for example, immediately after the handoff. Without that reference point, the same task can appear open in one collection and completed in another, producing a misleading comparison. The proposed formulas are in the metrics table.

Also track duration, interruptions by reason, and the team's perception of clarity. Shorter duration on its own may mean omission; more entries may mean better visibility. Separate missing fields, duplicate requests, external blockers, and legitimate changes to the plan. Do not use the raw count of open items to compare professionals with different workloads and acuity. Process metrics help investigate the work, but do not, on their own, demonstrate an effect on clinical outcomes.

Worked example: where follow-up gets lost

The chart shows an example classification of 40 open items at shift change: 18 with an owner and a deadline, 9 without an owner, 7 without a deadline, and 6 without either. The categories are mutually exclusive. In this example, 22 of 40 items, or 55%, have at least one of these two gaps. This does not mean that 55% of care was omitted, since the sample describes only the quality of the recorded information.

Use this kind of breakdown to choose the first change. If missing owners predominate, review assignment and confirmation; if missing deadlines predominate, examine how they are agreed upon. Repeat the collection using the same definition and check whether better documentation corresponds to real follow-up.

How to pilot in one unit before scaling up

Choose one unit and one frequent type of open item. Agree on the minimum fields with those who perform and those who receive the work, train on ambiguous situations, and do a baseline observation before the change. During the pilot, review a few items daily to check whether completion and confirmation mean the same thing to everyone. Scale up only when the workflow can represent exceptions without relying on additional informal instructions.

The next step is a joint review by nursing leadership, quality, and the teams involved: did the guide improve understanding, did the list duplicate documentation, and which blockers remain outside the unit's control? Staffing shortages, integration gaps, or another department's capacity cannot be solved with more fields. The expected output of this review is a specific process adjustment and a follow-up check, not a generic promise of error-free shifts.

Implementation checklist

• Define the changes, risks, and open items for the summary.
• Link open actions to the clinical record.
• Assign an owner and agree on a deadline or reassessment.
• Ensure the receiver can ask questions and confirm.
• Distinguish non-urgent requests from immediate escalation.
• Define completion, blocked status, and escalation.
• Audit a sample with stable criteria and share findings.

Tracking indicators

Handoff completeness (%)

100 × handoffs with all minimum fields ÷ audited handoffs.. Fix the fields and audit unit. Completing fields does not prove understanding.

Open items with complete assignment (%)

100 × open items with an owner and a deadline or reassessment ÷ open items at handoff.. Zero denominator: not applicable. Separate owner gaps from deadline gaps.

Overdue open items (%)

100 × overdue open items ÷ open items with a deadline, at cutoff.. Show items without a deadline separately. Investigate blockers and legitimate changes.

Interruptions during handoff

Observed interruptions ÷ observed handoffs, by reason.. Keep the observation unit constant. Urgent alerts may be necessary.

Process map: shift handoff and continuity

1. Prepare: Outgoing team reconciles documentation and actions.
2. Summarize: Sender highlights changes and priorities.
3. Clarify: Receiver asks questions and summarizes priorities.
4. Confirm: Owners agree on follow-up and deadlines.
5. Follow up: Incoming team reviews dependencies.
6. Reconcile: Record completions; reassess blockers and open items.

Assignment quality in 40 open items

Mutually exclusive classification of 40 open items. 22 have some gap; this does not measure omission of care.

• With owner and deadline: 18 open items
• No owner, with deadline: 9 open items
• With owner, no deadline: 7 open items
• No owner and no deadline: 6 open items

Example to illustrate the method.

Frequently asked questions

Is SBAR enough for a complete shift handoff?

It helps organize communication; responsibilities, confirmation, and follow-up are still needed. AHRQ presents SBAR and handoff as related tools. [1][2]

Does the open-items list replace the health record?

No. Formal documentation of the Nursing Process belongs in the health record, in line with the responsibilities set out in COFEN Resolution No. 736/2024. [3]

How should we handle an open item that depends on another department?

Identify who performs it and who follows up. Record the blocker, reassessment, and escalation, while respecting professional scope of practice.

Sources

1. TeamSTEPPS: Tool: Handoff
2. TeamSTEPPS: Tool: SBAR
3. Resolução Cofen nº 736, de 17 de janeiro de 2024

Scattered documents, outdated versions, informal onboarding, and training that does not reach every shift are distinct problems and need to be checked at the point of work. The proposed approach connects document governance and competency-based training. It starts with small, verifiable decisions: which version is valid, who the change matters to, how to teach it, and what evidence allows learning to be considered sufficient.

Separate documentation requirements from learning design

RDC 63/2011, an Anvisa regulation on good operating practices for health services, requires written, up-to-date procedures to be accessible to staff. It also requires training before staff begin their activities and on an ongoing basis, with records of date, time, hours, content, instructor, and workers involved. These are operating requirements that must show up in daily routine. [1]

This article proposes methods for putting them into practice, without presenting every method as a legal requirement. A single library, change history, and a competency matrix are management choices. Competency assessment must respect professional roles and the service's technical criteria. Neither an online course nor a signature expands the scope of practice of those who took part.

Start with an inventory and the risk of mismatch

List the documents used in a unit: protocols, standard operating procedures, workflows, forms, and quick reference guides. Record subject, target role, owner, approval, version, and where it is accessed. Look for duplicates by content, since files with different names may guide the same activity. Initial priority should consider the effect of conflicting instructions and frequency of use, rather than relying only on the age of the file.

A long list of expired documents can generate work without fixing the most critical issue. Start by observing a concrete task: does the professional know how to find the applicable guidance? Can they identify the version? Can they find a contact for questions? This test separates missing content, difficulty searching, and instructions that are hard to understand. Each cause calls for a different intervention and a defined owner.

Get the current version to the point of use

For each document, consider explicit states: drafting, review, approval, current, and superseded. Define who validates the content and who authorizes publication. A change must show its effective date, a summary of what changed, and the affected audience. The review interval should be chosen based on risk and relevant changes; a universal annual review is not established here as a regulatory requirement.

Use a stable link to open the current version and keep the history restricted to relevant purposes. When a quick reference guide summarizes a protocol, link the two and review both in the same change. Printed copies need identification and a replacement routine. When the system is unavailable, the service must know which controlled copy to use and how to reconcile it afterward, without promoting an old file to a reference out of convenience.

Test lookup using the language of frontline staff

A library organized by administrative structure can still be hard to use. Include terms the team actually uses, synonyms, and filters by activity or department. Check readability on the devices and in the locations available during the shift. A quick-access code helps only if the destination is correct, the device can be used in that setting, and there is an alternative when the network fails.

As a local exercise, ask professionals from different roles and shifts to find a piece of guidance in a simulated situation, with no patient data. Time the search and observe the path chosen, including searches that end on the wrong document. No universal maximum time is proposed. The result is used to compare the same task before and after adjustments, under similar complexity and conditions.

Turn reading into observable behavior

The WHO multi-professional patient safety curriculum guide offers a framework for planning teaching and supporting continuing education across different professions. [2] In the proposed management application, the starting point is to write down what someone should be able to demonstrate after training, instead of listing only the topics covered.

A simple matrix links role, activity, expected competency, reference version, assessment method, and current status. Depending on the risk, a question may check understanding, while a demonstration or simulation allows execution and communication to be examined. Completion should record the criterion used and who assessed it. If there is a gap, define supervision and a new learning opportunity, and avoid treating public exposure of results as an educational strategy.

Plan training around the real schedule

Organize the audience by role, unit, and shift, including temporary and contracted staff who perform the activity. Managers need to plan clinical coverage and participation windows. Asynchronous content can prepare the discussion, but does not by itself address hands-on skills. Duration and format should match the objective; short modules are an option, not a guarantee of learning or a solution to staffing shortages.

Assign facilitators for shifts with less access to the education team and establish how they will deliver the same guidance and the same criteria. Plan make-up sessions for vacations, leaves of absence, and hires between cohorts. Anyone who did not participate remains identifiable as pending, without dropping out of the denominator due to scheduling difficulties. The plan should state what can be done under supervision and who makes that decision in the service.

A new version requires an analysis of its educational impact

When a protocol changes, compare versions and classify the impact on the activity. Wording adjustments may call for a brief communication; a new work sequence or responsibility may require a demonstration and reassessment. These criteria are local proposals and need technical validation. The effective date should consider whether the people, supplies, and access needed to apply the guidance are ready.

The WHO surgical safety checklist implementation manual highlights education, multidisciplinary participation, leadership, and continuous feedback as implementation factors. [3] The lesson for protocol management is to involve those who perform the work before and after the change. Consulting night-shift staff helps identify constraints that a daytime administrative meeting may miss, without assuming that the same solution works in every context.

Read coverage without hiding gaps

Worked example: a unit needs to train 60 professionals on an updated procedure, with 20 eligible on each shift. By the implementation deadline, 18 completed the hands-on assessment on the morning shift, 16 on the afternoon shift, and ten on the night shift. The total is 44 of 60, or 73.3%. The chart shows these counts. The rates by shift are 90%, 80%, and 50%.

The difference calls for investigating access, schedule coverage, and assessor availability. It does not prove lack of interest or lower ability in the night group. Also check whether completion meant only attending or actually demonstrating the defined criteria. To compare different competencies, keep results separate: an aggregate total can mix tasks and levels of difficulty that are not equivalent.

Close the loop with observation and review

After implementation, review samples of use and collect questions. A nonconformity may stem from a hard-to-use document, a missing resource, an ambiguous criterion, or a learning gap. Do not prescribe more training before understanding the barrier. Also, do not automatically attribute an improvement in a clinical metric to the training: changes in staff, demand, and process may have occurred in the same period.

To get started, select a frequently used protocol, remove duplicates from the lookup path, publish the controlled reference, and test the search on every shift. Then apply an assessment proportional to the risk and keep open items with a deadline and an owner. The verifiable result is a team that finds the correct guidance, demonstrates what was defined, and has a path for questions and updates.

Implementation checklist

• Inventory documents and identify the owner and audience of each one.
• Record approval, current version, and effective date.
• Test the lookup link and the alternative for downtime.
• Document observable objectives and assessment criteria by role.
• Map eligible staff by shift and plan coverage for participation.
• Record training with the required fields and evidence of the assessment used.
• Schedule make-up sessions and reassessment of gaps with an owner and a deadline.
• Check copies at the point of use after an update.

Tracking indicators

Access to the correct guidance

100 × simulated searches ending on the current, applicable document / simulated searches performed, stratified by task and shift.. Also record search time and failures. Compare similar scenarios; quickly finding the wrong document is not a success.

Assessment coverage by shift

100 × eligible professionals who completed the required assessment by the deadline / eligible professionals on that shift at the defined point in time.. Document staff movements, leaves, and replacements. Completing the assessment does not imply passing it.

Demonstrated competency

100 × assessed professionals who met all mandatory criteria / professionals assessed for the same competency and version.. Present alongside coverage and the number pending; a high pass rate with low participation can hide gaps.

Compliance of copies in use

100 × audited access points with the current, applicable version / access points audited in the period.. Include physical and digital media used in daily routine. Describe the sample and do not extrapolate to locations not observed.

Process map: from protocol to on-shift competency

1. Identify the need: Quality and the team analyze risk, questions, and changes in the work.
2. Review and approve: Owner and technical leads validate content, version, and applicability.
3. Prepare implementation: Education and managers define competencies, eligible staff, resources, and coverage by shift.
4. Publish and replace: Document management releases the current link and controls previous copies.
5. Teach and assess: Facilitators record training and check criteria. Gaps go back to learning with defined supervision.
6. Observe and update: The team and quality analyze real-world use and send questions back to the document owner.

Hands-on assessment completed by shift in one unit

Implementation deadline with 20 eligible professionals per shift: morning 18/20, afternoon 16/20, and night 10/20. Total 44/60. Completion does not mean passing.

• Morning, 20 eligible: 18 professionals with completed assessment
• Afternoon, 20 eligible: 16 professionals with completed assessment
• Night, 20 eligible: 10 professionals with completed assessment

Example to illustrate the method.

Frequently asked questions

Does signing that you read the protocol prove competency?

The signature records acknowledgment under the terms defined by the service. To demonstrate competency, you need an assessment criterion consistent with the activity and its risk, respecting professional roles.

Do all protocols need an annual review?

Article 51 of RDC 63/2011 requires updating and access, without setting a universal annual interval in that provision. Define reviews based on risk and relevant events, checking any specific applicable regulations. [1]

How can we include the night team without repeating a long class?

Plan facilitators, coverage, asynchronous preparation when useful, and hands-on assessment at accessible times. The format can vary, but the criteria for the same competency must be consistent.

Sources

1. RDC 63/2011: boas práticas de funcionamento para serviços de saúde
2. Patient safety curriculum guide: multi-professional edition
3. Implementation manual WHO surgical safety checklist 2009

This article focuses on the operational organization of the sterile processing department (SPD), known in Brazil as the Central de Material e Esterilização (CME). The proposed registry, metrics, and pilot are management analyses to be validated with the technical lead and the departments involved. No times, temperatures, pressures, concentrations, or other sterilization parameters are provided, nor instructions for selecting or changing cycles.

RDC 15 and the regulatory review: separate requirements from proposals

In the official text consulted on October 26, 2026, articles 25 and 26 of RDC 15/2012, the Anvisa regulation on processing of health products, establish lot-level documentation for Class II SPDs and contract reprocessing companies and allow either a manual or an automated system. Retention follows specific legislation or, in its absence, a minimum of five years. The service's classification should take article 3 into account. [1]

Anvisa approved proposals for public consultation in September 2026. The Brasil Participativo portal lists CPs 1.420 to 1.423, related to health services, reprocessing companies, quality assurance, and device classification. [2][3] These are proposals, not automatically effective replacements. In the sources consulted, no final act replacing RDC 15 was identified. Before revising institutional procedures, check how these processes evolve and the applicable health regulatory guidance.

Choose what needs its own identity

As an operational design, distinguish the individual instrument, the set composition, and the package prepared for a given processing run. The name minor surgery tray can identify a type of set, but it does not distinguish two trays in use at the same time or their compositions at different moments. The registry must make it possible to answer which set that was and which components were associated with that preparation.

Start with the level of detail needed to answer the service's questions. Individual identification may be appropriate for certain instruments; in other cases, traceability starts with sets and packages. Do not promise to locate a specific piece if the system only records the set. Also preserve the history of replacements: updating the current record must not rewrite the composition of a past processing run.

Design the handoffs and their proof

For each step, consider an identified input, a verifiable output, and a person or role responsible for the record. The WHO and PAHO manual treats reprocessing as work that depends on coordinated infrastructure and procedures, providing a technical reference that complements Brazilian legislation. [4] The map below summarizes transfers of responsibility and does not function as a processing protocol.

At intake, link the origin to the set received and record discrepancies. At preparation, associate the verified composition with the package. During processing, preserve the link between packages and the recorded run. At distribution, keep the confirmed destination. A shift change should transfer pending documentation items with a defined owner, so the next team does not have to reconstruct decisions from loose sheets of paper.

Link evidence to the release decision

The proposed registry needs to separate completion of a step, available result, and release decision. A completed cycle record does not by itself show that all relevant checks were evaluated. It is up to the responsible professionals to apply the current technical and institutional criteria. The system should make pending items visible and record who decided, when, and based on which evidence, without replacing that assessment.

The label is also part of the chain: articles 83 to 85 of RDC 15 cover identification, legibility, and label information, including product, lot, relevant dates, method, and person responsible for preparation. [1] As a system proposal, maintain a searchable identifier that connects the package to internal evidence. Test scanning and lookup under real transport and delivery conditions, without changing technical packaging requirements.

Close out distribution and investigate returns

Distinguish release from sterile storage, receipt by the unit, opening or use, and return unused. Delivery to the operating room does not prove use in a procedure. To link the package to its use, the responsible team needs to record that event in the appropriate clinical workflow. This integration is an operational design proposal; its scope must be agreed upon and cannot be assumed from a scan at dispatch.

If there is a link to the health record, use institutional identifiers and access consistent with job responsibilities. Management dashboards can show status and pending items without exposing patient identification. When a package is returned, record its origin and reason before the routing defined by the responsible person. Avoid simply returning it to available inventory because the physical count reappeared: its history and condition need to be evaluated.

Make exceptions show up as pending work

An illegible label, a duplicate identifier, a mismatched composition, or a missing documentation link should open a visible pending item. Define who assesses the case and what evidence allows it to be closed. A free-text field that accepts any justification can make administrative closure easy while leaving the root cause in place. On the other hand, a hold with no owner can turn the system into a queue that no one can clear.

In the proposed design, each correction keeps the previous value, new value, reason, author, and time. The goal is to allow later review. During downtime, use forms with controlled identification and defined reconciliation, preserving both the time of the event and the time of later entry. This makes it possible to distinguish data entry delays from a real lack of information, without pretending that the system was available when the event occurred.

Use the chart to test how failures are classified

The chart contains an example of 30 documentation issues in one month: 12 with no confirmation of receipt at the destination, 8 with an illegible label, 6 with a mismatched composition, and 4 with no link between the package and the processing run. Each issue has one primary category.

In this example, the largest category would point to an investigation of the handoff between dispatch and receipt, not an automatic conclusion about processing quality. It would be necessary to observe the work and check for possible records in another system. The total number of documentation failures is also not an infection rate or a direct measure of sterility. Use the example to practice questions and criteria, without treating it as a performance benchmark.

Combine completeness, retrieval, and responsiveness

Completeness should evaluate a predefined chain, with origin, identifiers, evidence, decision, and destination. Counting completed fields is not enough. Take a sample and try to reconstruct the path without relying on anyone's memory. Record the time and the gaps. In another test, start from a package and find the associated records; then start from a processing run and locate the corresponding destinations.

Also track documentation rework and returns, always with stable denominators and criteria. A better rate may result from improperly excluding exceptions or from an easier sample. That is why checks should be spread across shifts and include different types of sets. The metrics proposed below evaluate information management; improvement in them does not by itself demonstrate clinical safety or full regulatory compliance.

Run a pilot that includes failures and outsourcing

Choose one family of sets and follow a complete path, with participation from the SPD and the receiving unit. Observe intake, preparation, records, distribution, and return. Include an unreadable label, an incomplete set, and system downtime in controlled exercises. Use fictional data in the exercises and keep a clear separation between training and real operational records.

If there is a contract reprocessing company, test whether the identifiers and documents received make it possible to reconstruct the path between organizations. Availability of an external portal does not guarantee future access to the history. Agree on export, correction responsibilities, and record retrieval. Expand the scope only after proving lookups in both directions and handling of pending items, with technical validation. Software can support this chain, but it does not by itself provide physical capacity, staff, or professional judgment.

Implementation checklist

• Confirm the service's classification and scope and record the regulatory version used in the review.
• Define the identity of set, package, and processing run and test for duplicates.
• Preserve historical composition when an instrument is replaced.
• Verify that looking up a package retrieves the associated records and release decision.
• Record destination confirmation and distinguish distribution from use.
• Test downtime procedures and reconciliation while keeping actual times and authorship.
• Run lookups in both directions and document gaps and who is responsible for correction.
• Review the 2026 public consultations before updating protocols for any future regulation.

Tracking indicators

Complete documentation chains

Sampled packages with all links and evidence defined in the audit protocol / total packages sampled × 100.. The sample should include shifts and exceptions; a completed field without evidence does not meet the criterion.

Time to reconstruct the path

Median minutes between the lookup request and retrieval of the required links in documented exercises.. Report the number of exercises and incomplete searches; use comparable tasks.

Distributions without confirmed receipt

Deliveries whose local confirmation deadline passed with no destination record / total deliveries with an expired deadline in the period × 100.. A proposed logistics control with no universal deadline; do not confuse lack of confirmation with proven loss.

Documentation rework

Processing runs with at least one later documentation correction / total processing runs in the period × 100.. Classify the reason and preserve versions; a drop may reflect fewer errors being identified.

Process map: documentation chain in the SPD

1. Receive and identify: Intake links origin, set, and discrepancies.
2. Record steps prior to preparation: The performing team documents steps according to applicable technical procedures.
3. Inspect and prepare: Preparation preserves composition, package identity, and required identification.
4. Link to processing: The team links packages to the run and the corresponding evidence.
5. Assess and record release: The responsible professional documents the decision and pending items according to applicable criteria.
6. Store and distribute: Sterile storage records release; destination confirms receipt.
7. Record use or return: The receiving unit records the applicable event and keeps a searchable link.
8. Review and investigate: SPD and quality retrieve the path in both directions and address gaps.

Documentation issues in one SPD month

Thirty issues, each classified by one primary cause. Absolute count of documentation issues; does not measure contamination, sterility, or clinical incidence.

• Receipt at destination not confirmed: 12 issues
• Illegible label: 8 issues
• Mismatched composition: 6 issues
• Package not linked to processing: 4 issues

Example to illustrate the method.

Frequently asked questions

Does RDC 15 require purchasing specific software?

Article 26 allows either a manual or an automated system. [1] The choice must demonstrate record retrieval and consistency, and work under the service's real conditions.

Does a code on the package prove sterility?

No. It is an identification tool. Technical assessment depends on the process and the relevant evidence. The code must allow the history to be looked up, without being treated as a self-sufficient certificate.

Have the 2026 public consultations already replaced RDC 15?

The sources consulted present regulatory proposals. [2][3] A public consultation does not make its draft a final regulation in force; follow any publication and transition rules before changing procedures.

Sources

1. RDC nº 15, de 15 de março de 2012: texto no AnvisaLegis
2. Processos deliberados na 17ª Reunião Ordinária Pública de setembro de 2026
3. Consultas públicas da Anvisa: CPs 1.420, 1.421, 1.422 e 1.423, de 23/09/2026
4. Decontamination and reprocessing of medical devices for health-care facilities, 2016

This article examines two operational pain points: time-consuming, duplicate documentation, and incomplete or denied claims. The proposal is a management method to test locally, without estimating savings in advance. Documentation quality should support continuity of care and make it possible to understand the care provided. Whether a service can be billed also depends on authorization, contract, coding, and other requirements; adequate documentation does not guarantee payment.

What must remain in the record

COFEN Resolution 736/2024 (from Brazil's Federal Nursing Council) requires formal documentation of the Nursing Process in the paper or electronic health record and distinguishes the responsibilities of registered nurses from those of other team members. Nursing diagnosis and nursing orders are exclusive responsibilities of the registered nurse. Simplifying screens requires preserving this division of responsibilities, without turning all documentation into a generic checkbox. [1]

COFEN Resolution 754/2024 covers records and digital systems and revoked Resolution 429/2012. The Code of Ethics, approved by Resolution 564/2017, requires the information needed for care to be clear, complete, and in chronological order. [2][3] When reviewing a form, the useful question is which clinical decision each field supports. An additional financial field needs its own purpose and owner; it must not alter the true description of the care provided.

How to locate documentation rework

Start by following a sample of claims from the encounter through the first billing review. Include accepted and returned claims, different shifts, and different types of encounter. Define the period and selection criteria before looking at the results. A sample made up only of problem claims reveals failures, but does not allow you to estimate how often those failures occur across total volume.

For each repeated piece of information, record its origin, destination, the professional involved, approximate time, and the stated reason for the repetition. Distinguish unnecessary copying from a review with a clinical purpose. Confirming a current condition is not the same as transcribing an administrative data point. If a spreadsheet receives information from the health record because another department cannot access it, investigate permissions and available reports before building an integration. The solution may lie in how access and data ownership are organized.

Classify the cause before assigning responsibility

A denied claim may involve missing evidence of care, a quantity mismatch, an incompatible authorization, a contract rule, or a registration error. These classes are an analysis taxonomy proposed here, to be adapted to each service. Avoid logging everything as a nursing failure. The reason received from the payer should be preserved and linked to the internal root-cause hypothesis, which requires documentary verification.

In Brazil's private health insurance sector, the TISS Standard organizes the exchange of information among participants. Consult the components published by ANS, the national private health insurance regulator, and the version that applies to the submission. [4] Separate the code received from the payer, the audit analysis, the person responsible for action, and the outcome of the appeal. A denial initially attributed to documentation may end up recognized as a contract discrepancy; keep the history of that review so you do not train the team on the wrong cause.

A workflow with prevention and feedback to the source

The proposed process starts with care and its documentation, moves on to reconciling claim line items, and ends with structured feedback on discrepancies. The clinical team documents what actually happened. Audit and billing check consistency, supporting documentation, and applicable requirements. Each open issue receives an owner, an internal deadline, and a resolution criterion. The map below represents this sequence and should be adjusted to the roles that exist in the institution.

When a discrepancy is found before submission, the feedback must state the problem precisely: which event, which piece of information, and which evidence is available. Vague requests to complete the chart generate more back-and-forth. A later clarification must follow the institutional correction procedure, preserve authorship, and distinguish the time of the event from the time of documentation. No one should invent care that was not performed or retroactively change the account to make the claim appear eligible.

Integrate data without multiplying errors

A useful integration carries data with context: encounter identifier, origin, time, author, and validation status. In the operational design, define which system is responsible for each piece of information and how an update reaches the others. Test prescription changes, cancellations, returns, and system downtime. An integration that only works on the usual path can generate duplicate charges precisely in the exceptions.

Structured forms help locate information, but too many mandatory fields encourage automatic answers. Test usability with professionals from different shifts and leave room for relevant specifics. Reusing registration data does not mean copying a previous assessment as if it were current. Resolution 754/2024 sets requirements for electronic records and signatures; the possibility of going paperless requires analyzing the conditions set out in the regulation and does not follow simply from having a login. [2]

Measure quality, rework, and financial impact separately

Track one documentation process measure, one effort measure, and one financial outcome measure. The first shows the share of claims returned due to a confirmed documentation issue. The second records minutes of rework per audited claim. The third tracks amounts denied and recovered by billing period. Publish numerators, denominators, and the follow-up window; recent claims may not yet have been reviewed or had an appeal decided.

The chart presents an example: of one hundred claims reviewed, forty had some issue, each classified by a single primary cause. Eighteen are documentation, ten authorization, seven contract, and five registration. The other sixty had no issue identified. These numbers show how to segment the problem. If documentation improves, other causes may continue to produce denials. A reduction in internal returns also does not automatically equal recovered revenue.

How to run a pilot and interpret its limits

Choose one unit and one claim type with enough volume for observation. Record the baseline, revise one form or one specific duplication, and track the same metrics afterward. Agree on who reviews questions and who decides on process changes. Compare periods with similar case mix and billing rules, noting any staff, contract, or system changes that could explain differences in results.

The expected result of the pilot is evidence for a decision: keep, adjust, or abandon the change. Check whether rework went down without losing necessary information, delaying care, or piling tasks up at the end of the shift. Operational analysis does not replace clinical judgment, technical audit, or contract review. The concrete next step is to choose one recurring issue, reconstruct its path, and test a fix at the source with nursing, audit, and billing.

Implementation checklist

• Define the sample, period, claim types, and inclusion criteria before the audit.
• Inventory repeated information with its origin, destination, purpose, and owner.
• Review local templates against COFEN Resolutions 736/2024 and 754/2024.
• Preserve the denial reason received and verify the internal cause before classifying it.
• Assign an owner and closure criterion to each open issue.
• Test cancellation, return, and correction before expanding an integration.
• Compare documentation quality, rework time, and financial outcome over equivalent windows.

Tracking indicators

Internal documentation return rate

Unique claims returned due to a confirmed documentation issue / unique claims audited in the period × 100.. Internal metric; not equivalent to the payer's denial rate. Report sample size and selection.

Documentation rework per claim

Sum of observed minutes spent locating, reconciling, or correcting information / audited claims, including those with no rework.. Record the observation method and separate active time from waiting time; do not use for individual rankings.

Initial documentation denials by value

Amount initially denied with a confirmed documentation cause / amount billed for the same claims in the billing period × 100.. Report the maturity of the billing period and attribution rules when there are multiple causes.

Documentation denial recovery

Amount actually received after appeal / documentation-related amount appealed in the same cohort × 100.. Track only cohorts with a stated decision window; an approved appeal without payment is not yet cash recovery.

Process map: from documentation to discrepancy analysis

1. Provide care and document: Clinical team: contemporaneous account and attribution of authorship.
2. Reconcile the claim: Audit and billing: check evidence, line items, and applicable requirements.
3. Resolve open issues: Originating department: clarify the discrepancy with traceability; return to reconciliation.
4. Submit and follow up: Billing: transmit the claim and keep the submission record and response.
5. Analyze denial or payment: Audit and finance: validate the reason, appeal, and payment received.
6. Fix the source: Department managers: revise the form or rule and monitor recurrence.

Primary causes of open issues in a claims review

Example of 100 claims reviewed: 40 with an issue and 60 with no issue identified. Each of the 40 receives only one primary cause. The values do not represent denials actually applied.

• Documentation: 18 claims with an issue
• Authorization: 10 claims with an issue
• Contract: 7 claims with an issue
• Registration: 5 claims with an issue

Example to illustrate the method.

Frequently asked questions

Does improving nursing documentation eliminate claim denials?

It can reduce documentation-related issues, but denials also involve authorization, registration, and contract rules. The contribution must be measured by confirmed cause and for the same cohort of claims.

Is it possible to eliminate all spreadsheets and paper forms?

It depends on the purpose of each document and on the system's technical and regulatory conditions. COFEN Resolution 754/2024 sets specific requirements; scanning or using a password is not enough to assume that any record can skip printing. [2]

What should we do first when there is a lot of duplication?

Choose one repeated piece of information, follow its path, and check why each department needs it. Remove a transcription only after confirming access to the source, context, and responsibility for updates.

Sources

1. Resolução Cofen nº 736, de 17 de janeiro de 2024
2. Resolução Cofen nº 754, de 16 de maio de 2024
3. Resolução Cofen nº 564/2017: Código de Ética
4. TISS: Padrão para Troca de Informação de Saúde Suplementar