Data Snapshot: Who goes to the doctor?
New data show that the largest divide is not along sex/gender, race/ethnicity, or even age-group lines.
Every year, the CDC conducts the National Health Interview Survey (NHIS). They’ve been doing this since 1957. The questions asked are varied, ranging from information about smoking habits to cancer history.
One series of questions focuses on healthcare use. For our Data Snapshot this week, we’re looking at the rates of people who responded that they had seen a doctor in the previous year, according to various demographics. (This includes doctors seen in hospitals, which includes ER and inpatient physicians. Dentists didn’t count.)
Here are the most recent results we have (from 2022):
Some of these findings are not too surprising. For example, 87% of females surveyed had seen a doctor in the last year, while only 78% of males had. Men are well-known for not loving to go to the doctor. The differences by race/ethnicity were present, though not as great as one might have expected; some of this could reflect the fact that seeing a doctor in the ER or a hospital counted the same as seeing one’s primary care doctor or specialist. Older data on primary care access show wider differences across these demographics. Also unsurprising, rates of having seen a doctor in the past year increased with older age.
However, the biggest line in the sand was not along sex/gender, race/ethnicity, or even age-group lines, but by insurance status. Those with insurance had a 34% higher rate of having seen a doctor in 2022 (that’s 86% versus 52%). I would have assumed that age would be the big dividing line. Nope. It’s insurance status. In my view, that 34% difference is pretty impressive. And it’s likely that those among the uninsured who did see a doctor did so during emergencies, rather than for preventative care or necessary screenings. This is not optimal.
How reliable are the datapoints?
The National Health Interview Survey has some real strengths. One is that in 2022, 27,651 people responded to the survey. That’s a lot of people. Another is that the survey is conducted in a consistent way, which makes it more likely that the answers are reliable. (How surveys are done matters a lot.) Another is that for many of the questions, we can track rates of certain responses from year to year to see how things have changed over time.
We’ve used the data from the NHIS survey here in Inside Medicine in the past. In fact, a post that Benjy Renton and I made on rates of immunesuppression in the US all the way back in March of 2023 eventually was used wholesale by some other researchers for a publication that appeared in the Journal of the American Medical Association earlier a few months ago. (Are we bitter about this? Maybe a little. But, we had other fish to fry.)
But the big weakness with respect to the NHIS survey is this: Who exactly fills these things out? The survey is long (with potentially up to 600 questions) and can be quite time-intensive to complete. If I got an invitation to participate in this survey, I’d politely decline. Indeed, it’s likely that survey participants are more likely than average to have complicated medical needs; people who have a vested interest in something tend to be more willing to be polled about that topic. (For example, 9.2% of respondents to the NHIS survey reported being hospitalized at least once in 2022; that’s around 26% higher than the rate among the general US population. The survey researchers do correct for the demographics of the participants to make sure that certain groups’ contributions are not over- or underweighted in the reported outcomes. But the researchers can’t control for who in general agrees to complete these surveys.
So, while I think the NHIS data carry important insights, it’s not a perfect reflection of the general US population. Still, it’s hard to deny that data on insurance status and doctor visits. And, if we want people to see their doctors before they get quite sick, shoring up rates of people with health insurance in this country would be a good approach. What’s the big innovation that improved rates of insurance in the US this century? The Affordable Care Act. Too bad that is constantly under attack.
Questions? Ideas? Insights? Chime in below in the Comments section.
Thanks to Benjy Renton for curating this Data Snapshot.
Id be very curious how this data was also affected by Covid. As 2021-2022 was the height of the Pandemic with excess deaths reaching peaks in both Feb 21 and Mar 22 . I would expect that medical reality of severe cases and fear of getting very ill drove attendance at both Primary Care and ER visits. We know that most hospitals ER's were drastically overwhelmed.....
Is there similar data from 2018?