4.1 Healthcare Statistics and Rate Formulas

Census, Service Days, and Occupancy

The inpatient census is the number of inpatients present in a facility at a given time, conventionally taken at midnight. The daily inpatient census adjusts the midnight census for patients who were both admitted and discharged the same day (intra-day patients), so they are still counted.

An inpatient service day (IPSD) is a unit of measure for the services received by one inpatient in one 24-hour period. The total inpatient service days for a period is essentially the sum of each day's daily census.

The average daily census is the mean number of inpatients present each day for a given period:

Average daily census = Total inpatient service days for a period / Total number of days in the period

Newborn (NB) service days are usually reported separately from adults and children (A&C), so be careful which population a question asks about. If a question gives 9,300 A&C service days for a 30-day month, the A&C average daily census is 9,300 / 30 = 310. A frequent distractor mixes newborn days into an A&C-only question; read whether the prompt says "adults and children," "newborns," or "total."

Two related counts the exam expects you to keep straight: a transfer within the facility is not a new admission, and a patient admitted and discharged the same day is still counted in the daily census even though that patient is not present at the midnight census.

Bed Occupancy and Length of Stay

The bed occupancy rate (percentage of occupancy) compares the service actually provided to the maximum service the bed complement could provide:

Bed occupancy rate = Total inpatient service days for a period / (Total bed count days in the period) x 100

where bed count days = number of beds available x number of days in the period. A common trap: when the bed count changes mid-period, you must use the actual bed-count-days, not a single bed number.

Length of stay (LOS) for one patient is the number of calendar days from admission to discharge; same-day admit/discharge counts as 1 day. The total length of stay (discharge days) sums the LOS of all discharged patients.

Average length of stay (ALOS) = Total length of stay (discharge days) / Total number of discharges (including deaths)

ALOS is built from discharged patients, not from the census — mixing the two is the classic error. For example, if 40 patients were discharged with a combined 280 days of stay, the ALOS is 280 / 40 = 7.0 days.

The median length of stay is often reported alongside ALOS because a few very long stays skew the mean upward; the median better represents the typical patient. Note the difference between discharge days (counted at discharge for ALOS) and inpatient service days (counted each day for census and occupancy): they are tabulated from different events, and substituting one for the other is a guaranteed wrong answer.

Mortality (Death) Rates

Mortality rates are expressed as percentages, calculated to three decimals and rounded to two. Patients dead on arrival (DOA) and outpatients who die in the emergency department are not counted as inpatient deaths, and fetal deaths are never included in hospital death rates.

• Gross (hospital) death rate = Total inpatient deaths (incl. NB) / Total discharges (incl. deaths) x 100.
• Net (institutional) death rate = (Inpatient deaths − deaths < 48 hrs) / (Total discharges − deaths < 48 hrs) x 100. It removes deaths within 48 hours of admission, on the theory the hospital had little chance to affect them, so the net rate is always ≤ the gross rate.

Worked example: 8 deaths, 2 of them under 48 hours, among 400 discharges. Gross = 8/400 x 100 = 2.00%. Net = (8−2)/(400−2) x 100 = 6/398 x 100 = 1.51%.

A further refinement, the postoperative death rate, is deaths within 10 days of surgery divided by the number of patients operated on. The anesthesia death rate divides deaths caused by anesthetic agents by the number of anesthetics administered. As with all mortality measures, the numerator and denominator must describe the same population — operative deaths over operative cases, anesthesia deaths over anesthetics. Because the 48-hour cut is removed from both halves of the net rate, the net rate can never exceed the gross rate; if your computed net rate comes out higher, you have made an arithmetic error.

Obstetric, Infection, Autopsy, and Consultation Rates

Obstetric and infection rates each have a specific denominator the exam expects you to know exactly:

Key distinctions: the C-section denominator is deliveries, not births. A nosocomial (hospital-acquired) infection must develop after admission; an infection present on admission (POA) is community-acquired and excluded. The net autopsy rate removes from the denominator the bodies released to a coroner or medical examiner (ME), because the hospital could not autopsy them.

Putting It Together

Most RHIT statistics items give you raw counts and ask for one rate. The fastest path is: (1) identify the population (A&C vs NB, deliveries vs births), (2) pick the matching numerator/denominator from the table, (3) multiply by the right base (100 for percentages, 1,000 for fetal/neonatal where specified), and (4) round to two decimals.

Watch the "including deaths" rule: a discharge denominator almost always includes patients who died, because a death is a form of discharge. The most-missed traps are forgetting that the C-section denominator is deliveries, that the net death rate strips the same 48-hour cases from both halves of the fraction, and that newborn data are usually tabulated separately from adults and children.