Supply of Medical Services

Metrics for Vaccination Rates

What is a good metric to measure if a doctor is adequately vaccinating their patients? For instance, is the pediatrician Dr. Smith doing an adequate job of giving his patients flu shots?

One of the easiest, and most common metrics used is simply the vaccination rate.

  • Vaccination rate = (nbr of patients vaccinated)/(total nbr of patients).

But does this truly capture whether or not Dr. Smith is doing a satisfactory job of immunizing his patients? Let us decompose the vaccination rate further.

  • Vaccination rate = P(Doc Visit)P(Vacc|Doc Visit) + P(No Doc Visit)P(Vacc|No Doc Visit).

For the time being, let us assume that people who do not go to the doctor do not get vaccinated [i.e.: P(Vacc|No Doc Visit)=0, so Vacc rate = P(Doc Visit)P(Vacc|Doc Visit) ]. In reality this assumption does not hold–because many people are immunized at pharmacies, Costco, their employer, their school, etc.–but it will make the analysis here simpler and does not change how physicians themselves should be evaluated.

We see that there are two components which determine the vaccination rate. The first is whether or not the patient visits the doctor. This probability is influenced by demographic factors, distance from the provider, education of the parents and many other factors. Dr. Smith can not change the demographics of his patient base or make his patients move closer to his office. Of course, items such as reminder phone calls may help increase the probability the patient visits his office, but many determinants of whether or not the patient will eventually reach his office in a given year remain out of his control.

On the other hand, the probability a patient is vaccinated once they arrive at the office is certainly under the control of the doctor. A better metric would be to analyze a group of patients eligible for flu vaccination and to see whether or not Dr. Smith actually administered the flu shot. 

A chart audit of each physicians would be useful to see if Dr. Smith vaccinated the patients who came to the office.  If the patient was not vaccinated, the auditor should look to find out why the patient was not vaccinated and work to improve the vaccination process. 

Using a chart audit to study the P(Vacc|Doc Visit), is a metric more relevant to doctors and can result in more concrete, more immediate improvements.  The downside is that is does not emphasize the public health aspect of trying to incentivize physicians offices to increase the P(Doc Visit), for their patient base.