Our research finds a correlation between patient satisfaction & COVID-19 deaths in U.S.
Here we present the methodology of our study showing that higher online review ratings of doctors indicate reduced percentage of COVID-19 deaths in the U.S. top 100 cities. Our model is also available as an interactive chart.
This research project used regression analysis on COVID-19 data from the Centers for Disease Control and Prevention (CDC),compiled per county by USA Facts, with the Happy Patient Index (HPI) data compiled by Vanguard. The data about COVID-19 deaths in the U.S. included the case count and death count as reported on October 25, 2020. The HPI measured how well patients rated their doctors online in the 100 most populous cities in the country.
COVID-19 data for the primary county of each of the 100 HPI cities was used to determine the respective COVID-19 mortality rate, which is the percentage of deaths from the total cases. Counties containing more than one HPI city were only counted once, leaving a total of 89 counties represented in the model.
The modeled data is accurate to the nationwide data, predicting 9.0 million infections (105% of actual), 206,000 deaths based on population (92% of actual), and 204,000 deaths based on cases (91% of actual). The model does not account for other factors that may affect COVID-19 mortality, such as: comorbidities, age, gender, race, etc. Existing data from the HPI showed cities had a patient satisfaction rating of 3.702★ on average.
Among many factors, patient satisfaction is substantially correlated with COVID-19 survivability. In every instance, areas with a COVID-19 mortality rate over 5% had a below-average patient satisfaction rating.
The five worst performers (deaths/cases rate, star rating) are the counties for:
- New York (8.94%, 3.53★).
- Detroit (7.68%, 3.57★).
- Jersey City, New Jersey (6.76%, 3.60★).
- Newark, New Jersey (5.61%, 3.61★).
- Buffalo, New York (5.05%, 3.44★).
All had patient satisfaction rates below the national average.
The five best performers (deaths/cases rate, star rating) are the counties for:
- Madison, Wisconsin (0.33%, 4.09★).
- Lincoln, Nebraska (0.49%, 3.79★).
- Anchorage, Alaska (0.59%, 3.86★).
- Lexington, Virginia (0.85%, 3.89★).
- Tulsa, Oklahoma (0.93%, 3.64★).
Except for Tulsa, all had patient satisfaction rates above the national average.
Two notable outliers (deaths/cases rate, star rating):
- New Orleans (4.41%, 3.97★) – surprisingly poor survivability despite high satisfaction ratings.
- Bakersfield, California (1.23%, 3.33★) – surprisingly good survivability despite low satisfaction ratings.
Satisfaction and Health Outcomes
Studies have found that patients’ positive experiences with the healthcare system result in better health outcomes. Our on-staff Medical Advisor Neil H. Baum, MD, says a patients-first philosophy is not only an attitude, it’s a strategy for a successful practice.
For patient ratings to be reliable, they must be genuine and without conflict of interest. Reviews must be nonburdensome, noncoercive and noncompensatory. Online reviews, such as those used in the HPI, have been shown to accurately reflect more traditional patient experience surveys, such as HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems).
The coefficient of determination for deaths/case, deaths/population, and cases/population equaled .0704, .1161 and .0652 respectively. The relatively low values can be expected due to the broad demographics and varied reaction to COVID-19. The CDC’s COVID-19 reporting is the most reliable data available, and the HPI is the only per-city measurement of patient satisfaction we are aware of.