Question

Think of an instance when the health administrator might use a hypothesis for quantitative analysis. State the hypothesis, null hypothesis, variables, and how the analysis would be done.

Answer 1

A health administrator work includes management and administration of public health systems, health care system and hospital networks. Besides financial aspects it also involves ensuring smooth operations. A quantitative analysis in the health administration segment can be done in any one or more, but not limited to, of the following cases:

• Number of different types of patient visit in a hospital,
• Number of patients visits in emergency room,
• Number of laboratory tests done in a day/week/month/quarter/year
• Number and type of surgical operations done in a certain period
• number and type of employees used to provide some specific services and cost involved in those services
• Daily volume of calls for ambulance
• demand vs supply of some specific services

Let's say the administrator is interested in testing if they have enough staffs for critical illness care service next 6 months.

Here the null hypothesis is H₀ : Number of staffs for the specific service is enough for next 6 months

i.e.,

To test the hypothesis we need to forecast the demand of critical illness service for next 6 months and then to see if the number of staffs available is significantly equal with the demand.

We'll do a time-series analysis to predict the future demand of the critical illness service.

Variables to be used: Historical data on critical illness service of at least 5 years, Number of request received for the service, number of service provided, number of days for which the services are provided, age of patient.

Using the above variables, we can run few mathematical models using Auto-regressive Integrated Moving Average (ARIMA) to predict the future demand and create a 95% or 90 % ( as per the requirement ) confidence interval around it.

using the above predicted demand, we will then estimate the number of staffs required using the standard practice of the health administrator, which is available for any industry. or it can be calculated using industry data.

Once we have a statistically satisfied model for predicting the future demand, we can move to the next steps which is analysis the number of staffs assigned to the critical illness service and to check if the number fall within the confidence interval created above.

Few points to note:

• there may be common staffs who provide critical illness service as well other services. Whole calculating the actual staffs for critical illness service, we need to keep in mind that what percentage of them can be used.
• if there any hiring plan in future, that should also keep in mind while counting the actual staffs.