How to measure OPD waiting time the NABH way: the formula, a register and a worked example

Most hospitals preparing for NABH can quote their OPD waiting time. Fewer can say exactly how it was measured, and that is where assessors look. This guide takes NABH’s own definition line by line, turns it into a register your front desk can start on Monday, works through an example, and lists the mistakes that quietly change the number.

What NABH’s OPD waiting time indicator says

The indicator is in the key performance indicator table at the back of the NABH Accreditation Standards for Hospitals, 6th edition, in force since 1 January 2025 (indicator 22, page 213 of the printed standard). It belongs to standard PSQ.3 in the chapter on Patient Safety and Quality Improvement: objective element c asks the hospital to identify and monitor key indicators for its managerial structures, processes and outcomes.

One phrase in the definition matters more than it looks: arrival at the OPD “may or may not be the same time as registration”. A patient who registers at a central counter at 8:40 and reaches the eye OPD at 9:05 starts waiting at 9:05.

NABH also gives a column of guidance for hospital software. The system should add up the waiting time of all outpatient consultations, and count the denominator up to midnight on the last day of the calendar month. Where software can’t work out the numerator or denominator, NABH’s introduction to the indicators asks that staff can at least type them in, so the value is still calculated automatically.

When does the clock start and stop?

It starts:

• for a walk-in, when the patient reaches the OPD where they will be seen;
• for a booked patient, at the appointment time, even if they came two hours early. Their extra wait isn’t counted, though it still fills your benches;
• for a booked patient seen before the appointment time, never: the wait is zero, not a negative number.

It stops when the consultant begins the assessment. A resident taking the history, a nurse checking blood pressure or a technician’s eye test before the consultant doesn’t stop it.

Grey areas the standard doesn’t settle. Decide each one once, write it into your indicator definition, and apply it every month:

1. A booked patient who arrives late. Read literally, the clock starts at the appointment time even if the patient came 30 minutes late. Our suggestion is to start at whichever is later, the appointment time or arrival, so the patient’s own lateness doesn’t count. Tell your assessor which rule you use.
2. Arrival windows (“come between 10:30 and 11”). Treat the start of the window as the appointment time, with the same “whichever is later” rule.
3. Patients who leave before being seen. They have no consultation, so they fall outside the formula. Count them separately: a rising count is a warning the average can’t show.
4. A second consultation the same day, such as a patient sent from medicine to cardiology. NABH’s software guidance adds up “all” outpatient consultations, so we would record one line per consultation, starting from arrival at the second OPD.

A register your front desk can keep

Paper is enough to start. You need two times per patient, written at the moment they happen, by the same clock.

• Reached OPD: written by whoever receives patients at that OPD, not at a registration counter elsewhere.
• Consultant began: written when the patient goes in to the consultant, by the doctor’s assistant or the doctor.
• Clock start and wait: worked out later by the quality team, using your written rules.
• Note: seen early, arrived late, resident saw first, left unseen, doctor called away.

Use one clock that every desk can see, or the registration computer’s time, never a mix of phones. Don’t fill the register from memory at the end of the session. At the end of each day, total the minutes and count the patients for each consultant; at the end of the month, add the daily totals.

A worked example

This is an illustration with made-up patients: one orthopaedic consultant’s Monday morning, with the OPD starting at 9:30.

Seven consultations waited 28 + 10 + 0 + 45 + 40 + 5 + 5 = 133 minutes in all, so the NABH figure for the morning is 133 ÷ 7 = 19.0 minutes. Measured from arrival instead, the same seven waited 223 minutes, an average of 31.9.

Three rows show why the rules must be written down:

• Token 4: stopping the clock when the resident began at 9:50 would have recorded 35 minutes instead of 45.
• Token 7: under the literal reading the wait is 30 minutes (10:00 to 10:30); under our rule it is 5.
• Token 8 never enters the formula. Only the “left unseen” count shows that someone gave up.

At the end of the month, add all minutes and all patients, then divide once. Don’t average the daily averages. If Monday’s 140 patients waited 7,700 minutes in all (55 on average) and Thursday’s 35 waited 525 (15 on average), the two daily averages give 35 minutes, but NABH’s formula gives 8,225 ÷ 175 = 47.0. The busy day has more patients, so it must weigh more.

Mistakes that quietly change the number

1. Using registration time as arrival. It hides a registration queue at the OPD’s own desk, and adds time if patients register early and report to the OPD later.
2. Starting booked patients at arrival, which counts their early arrival against you, or writing negative waits for patients seen early instead of zero.
3. Stopping the clock at the resident, nurse or technician.
4. Sampling, or dropping rows with a missing time. NABH says no sampling. Chase the missing time the same day, and if a row can’t be completed, count how many and report it. Standard PSQ.3.e expects the quality team to verify indicator data regularly.
5. Averaging daily or departmental averages instead of total minutes ÷ total patients.
6. Mixing clocks: a phone, a wall clock five minutes fast and the computer.
7. Mixing indicators. The wait for a blood test or X-ray is NABH’s separate indicator 23, waiting time for diagnostics, timed from the requisition reaching the counter until the test begins.
8. Changing the method mid-year, for new software or a new late-arrival rule, without a note. The trend breaks, and the assessor will ask why.

What NQAS asks of public hospitals

The National Quality Assurance Standards 2024 put waiting time under standard H4: the facility measures service quality indicators every month (measurable element H4.1) and works to improve them to meet benchmarks (H4.2). The standard’s intent names patient satisfaction scores, waiting time and LAMA rates.

The OPD checklist in the District Hospital assessor’s guidebook (2018, Volume I) lists under H4.1 the waiting time at the registration counter, the antenatal clinic, the general OPD, the paediatric clinic and the surgical clinic, plus the average door-to-drug time, all checked by record review. Consultation times sit under H3.1, with a time-motion study as the means of verification.

Unlike NABH, neither document gives a formula, and NQAS leaves benchmarks to each state. A public hospital working towards both can use NABH’s written definition for its general OPD figure and time the registration counter on its own.

How to use the number to improve it

Measurement is half of what NABH asks. Under PSQ.3, the hospital must also analyse the data to find opportunities for improvement (f) and implement and evaluate the improvements (g), and PSQ.4.a, a core element, expects quality improvement projects. In practice:

1. Split the figure by consultant, session, weekday and hour of arrival. A hospital-wide 30 minutes can hide one doctor’s 90-minute Monday.
2. Report the share waiting over an hour next to the average. NABH asks only for the average, and an average can hide a few very long waits.
3. Note when each session really started. If a 9:30 OPD starts at 10:15, every patient carries those 45 minutes.
4. Look at when patients arrive. At AIIMS New Delhi, 79% of OPD patients reported before 10 am and only 19% of them had been seen by then, which led it to stagger appointments. Bunched arrivals are a booking problem, not a doctor problem.
5. Test one change at a time. In a worked case in the government’s 2021 quality guidelines, a team timed 10 random patients a day, plotted them on a run chart, and tried each change idea for four to seven days before keeping, adjusting or dropping it. A sample is fine for testing; the NABH figure still counts everyone.
6. Share the number with staff every month, with what you are trying next.

For the fixes themselves, such as time windows, phone booking, registration before arrival and telling patients their real wait, see our guide to reducing OPD waiting time. For how patients feel about the wait, ask them with an OPD feedback form.

Where DrMate fits

DrMate’s front desk shows every patient of the day as booked, arrived or with the doctor, and its 7- and 30-day report includes an average wait. Before any software’s wait figure goes into your NABH file, DrMate’s included, check its definition against NABH’s: for booked patients the clock starts at the appointment time, not at arrival, and it stops only when the consultant begins. DrMate’s AI receptionist books callers into genuinely free times with the right doctor, so more of your patients have an appointment time to be measured against.