Data Vs. Knowledge: When Physician Specialists Are Away, Patient Survival Rates Go Up

Are a lot of doctors bad for your health?

Probably not, but the fact that their absence in certain critical situations provides better results than their presence suggests that there is still a lot we don't know about patient care.

This will not be a mysterious finding to anyone who has endured a hospital stay. But it is an especially interesting finding to reveal during the holidays when fear of what understaffing may do to their chances may cause people to delay seeking care around holidays out of fear that their chances of recovery are better if they can just hold on until things return to normal.

These results are robust, which means, statistically, that they have been 'normalized' to account for vagaries in almost all of the factors that might influence outcomes, like age, wealth, location, disease type, etc. In other words, this is not a fluke. But it may suggest that the way care is delivered in acute situations may need to be rethought. Which is why data is fine, but knowledge is better - and wisdom is best. JL

John Timmer reports in ars technica:

At teaching hospitals, the rate of death after heart failure was 24.8 percent on non-meeting days. While the cardiologists were out of town, it dropped to 17 percent. With cardiac arrests, where death rates fell from 68.6 percent to 59 percent while meetings were happening.
Don't get sick on a weekend." That advice is also part of a title of a research paper that evaluates the fates of patients who go through the emergency room on a weekend. These patients are more likely to die. It's just one of a number of studies that suggests patients who enter the hospital while the staffing is lower or the staff more relaxed end up with worse results.
But the precise cause of this enhanced weekend mortality has been hard to determine; is it the reduced staff, a more leisurely approach to care, or some other factor? To try to get at the cause, some researchers obtained records of heart patients who had a critical event during a time when hospitals were at full staff, but heart specialists were likely to be out of town. Unexpectedly, they found that the patients did significantly better when the relevant specialists were unavailable.
The study relied on medicare records to track patients that were admitted to a hospital with a serious heart condition: acute myocardial infarction, heart failure, or cardiac arrest. The key measure was simply whether the patient was still alive 30 days later.
That may sound simple, but the rest of the analysis was remarkably sophisticated. To figure out when heart specialists were most likely to be present at hospitals, they selected two large cardiology meetings: the American Heart Association and the American College of Cardiology, both of which attract over 10,000 participants. Patients admitted during the meetings were compared with groups admitted three weeks before and after. Reasoning that researchers are more likely to attend these meetings, they analyzed teaching hospitals separately from regular ones.
As additional controls, they checked a number of additional meetings for oncology, gastroenterology, and orthopedics specialists. They also looked at the impacts of additional critical injuries, like gastrointestinal bleeding and hip fractures, as well as non-critical cardiac problems.
In total, there were tens of thousands of patients involved. And the trends were clear. At teaching hospitals, the rate of death after heart failure was 24.8 percent on non-meeting days. While the cardiologists were out of town, it dropped to 17 percent. A similar trend was apparent with cardiac arrests, where death rates fell from 68.6 percent to 59 percent while cardiology meetings were happening. There was no significant difference with acute myocardial infarction patients.
So, having specialists in town appeared to make matters worse for patients—the exact opposite of the hypothesis the researchers set out to examine. The various controls suggested the effect was robust, and it persisted after adjusting for other potential influences, like age and sex.
In a press release accompanying the report, one of its authors, Anupam Jena, said "That's a tremendous reduction in mortality, better than most of the medical interventions that exist to treat these conditions." What could possibly be causing it? The authors consider three possibilities. First, there's something involved with the changes in cardiology staffing that occur when specialists go out of town that actually increases care. The second is that there are fewer people having outpatient or same-day procedures, given that doctors wouldn't schedule these when they knew they'd be absent. This would allow the remaining physicians to better focus care on the serious cases.
The final possibility that they consider is that the doctors that remain behind are more cautious about the care they give, avoiding aggressive procedures such as the use of angioplasty or stents to re-open clogged heart vessels. This would be consistent with the lack of effect in acute myocardial infarction patients, where this procedure is used less often.
Although their analysis can't distinguish among these possibilities, it's clear that this effect warrants further attention. Both because it's possible that the long-term survival evens out thanks to more aggressive treatment, and because we might find out that we've been acting a bit too aggressively.