Can medical records serve both statistical and patient needs?

Reader Jay, in a comment on my post about health-care costs tanking the economy, raises an interesting question about the sorts of standardized medical records that would be needed to evaluate efficacy (and therefore economic efficiency) of various treatments:

The idea is clearly to have standardized health-care records systems so that data can easily be aggregated and analyzed
[he writes, quoting my post.]

That's a shift in priorities, away from records centered around benefit to the patient, a subtle but not insignicant difference. Broad based statistical research involves, though necessity categorization, standardization and a significant degree of pigeon-holing. As a patient I would want my records organized around me and not convenience of government databanks.

Patient records are ideally almost the opposite of this. Now I guess both sorts of records can be run in parallel, but that seems like more wasted effort for doctors.

I see his point, and this is outside my bailiwick, but I would think records could serve both purposes. A good patient record for hospital and provider uses would clearly show what? -- diagnosis or diagnoses under treatment; related diagnoses; key relevant variables in patient health and demographics like age, etc; prior conditions; and treatments and outcomes, as well as ancillary stuff like what they ate and so on. It seems to me those same things are just what you want to measure when you evaluate whether a particular drug or device or treatment protocol works.

Or am I missing something? Is there a doctor in the house? Records wonk?

More like this

A common refrain among practitioners and advocates of alternative medicine is that the reason randomized clinical trials frequently fail to find any objective evidence of clinical efficacy for their favorite woo is because, in essence, science is not the right tool to evaluate whether it works. In…
The work up of "fever of unknown origin" (FUO) is a classic exercise in internal medicine. Originally defined as a temperature greater than 38.3°C (101°F) on several occasions for more than three weeks with no diagnosis after one week of inpatient study, the definition has shifted. This…
Check this very scary projection of what current trends in health-care spending will mean for our economy: a growing weight that will account for half of GDP by 2082: Peter Orszag, Obama's budget director, shows that slide in his standard talk on what's wrong with our budget. It shows why, as Ezra…
I advocate science-based medicine (SBM) on this blog. However, from time to time, consider it necessary to point out that SBM is not the same thing as turning medicine into a science. Rather, I argue that what we do as clinicians should be based in science. Contrary to what some might claim, this…

Patient records are ideally almost the opposite of this. Now I guess both sorts of records can be run in parallel, but that seems like more wasted effort for doctors.

As a non-doctor I am probably misunderstanding something - but if you're keeping records for two different purposes, I don't see why there must be a great deal of 'wasted effort'. Identify the common parts (being careful not to mistake similar parts for common parts) and copy and paste. If there are really no common parts, no effort is 'wasted' because the two sorts of records are quite different. If there are common parts (as seems likely), effort is conserved.


EHRs are very sophisticated. I can create templates to do anything, and i can allow access for data to be collected on anything.

sure...medicare is beginning to collect data under the "PQRI" program, tracking various disease-specific quality measures. currently, it's up to individual practitioners to report coded values to medicare. I will be able to create a template in my EHR to do that more or less automatically.

Of course, i may have misunderstood the topic at hand.

I guess revere would be the expert, but I thought medical records are routinely used in epidemiology. It's not my area (the nearest I got was plant epidemics, where patient confidentiality is less of an issue), so I don't know how it works in practice. I suspect it also varies by country.

My impression, Bob OH, is that the use and standardization of health records and treatment/outcome statistics does vary substantially among countries -- and that the US, with such a fragmented system, badly lags other countries in that aspect. It is the same way with our educational system, and with the same effect: The lack of standard metrics for methods and outcomes makes it extremely difficult to know what works and what doesn't, which of course makes it hard to eliminate practices that don't work and find overlooked practices that do work.

From an IT perspective, once you have standardized patient records, it's simple to have both patient and aggregated data. The patient data can be rolled up into a data warehouse, and sliced and diced appropriately. The technical solutions are well-known and available.

The problems are political rather than technical. In an insurance-driven healthcare system, I don't want my standardized records made available; since it's certain that data will be used to deny me coverage.
Further ranting here,