Behavioral Ecology Blog

Ask any 1st year Statistics major about the difference between causation and correlation, and chances are, they’ll nail it every time… Given the simplicity of the issue, and the widespread manner in which it’s taught, you’d expect the amount of confusion to approach 0% in folks with advanced degrees in the sciences- especially those doing basic scientific research. Now in my field- evolutionary biology, confusion risks embarresment, ridicule, or rejection, but in the medical sciences, confusion might mean peoples life.. It is for this reason that the difference between causation and correlation should be extremeley well noted on the medical literature.

Unfortunately, in a recent study published in JAMA (fulltext), and highlighted in ScienceDaily about hematocrit and surgical risk- the 2 relationships are seriously confused…

Results Thirty-day mortality and cardiac event rates increased monotonically, with either positive or negative deviations from normal hematocrit levels. We found a 1.6% (95% confidence interval, 1.1%-2.2%) increase in 30-day postoperative mortality associated with every percentage-point increase or decrease in the hematocrit value from the normal range. Additional analyses suggest that the adjusted risk of 30-day postoperative mortality and cardiac morbidity begins to rise when hematocrit levels decrease to less than 39% or exceed 51%.

Conclusions Even mild degrees of preoperative anemia or polycythemia were associated with an increased risk of 30-day postoperative mortality and cardiac events in older, mostly male veterans undergoing major noncardiac surgery. Future studies should determine whether these findings are reproducible in other populations and if preoperative management of anemia or polycythemia decreases the risk of postoperative mortality.

There is a similar story in the American Journal of Cardiology, Prognostic Value of Admission Hemoglobin Levels in ST-Segment Elevation Myocardial Infarctionnext term Patients Presenting With previous termCardiogenic Shocknext term (full text)

Even in the era of primary percutaneous coronary intervention (PCI) in patients with ST-segment elevation myocardial infarction (STEMI) complicated by cardiogenic shock (CS), mortality remains high. Whether admission hemoglobin (Hb) concentration is a predictor of mortality in patients with CS treated with primary PCI is unexplored. We assessed the relation between admission Hb concentration and 1-year mortality in patients with STEMI and CS who were treated with PCI at admission. We investigated a cohort of 265 patients with STEMI with CS on admission. Patients were categorized in 3 groups according to plasma Hb levels at admission: 9.6 g/dl (group I, n = 22), 9.6 to 12 g/dl (group II, n = 59), and >12 g/dl (group III, n = 184). All-cause mortality at 1 year was 64%, 46%, and 35% for groups I, II, and III, respectively (p = 0.007). Multivariate logistic regression analysis showed that the odds for mortality increased 17% for every 1.0 g/dl decrease in plasma Hb (odds ratio 1.17, 95% confidence interval 1.01 to 1.35, p = 0.042). In conclusion, admission Hb concentration is an independent predictor for 1-year mortality in patients with STEMI undergoing primary PCI.

Do you see the problem here??? Let’s start by making a list of the medical conditions that are known to be associated with anemia.

1. nutritional deficiencies
2. kidney disease
3. certain cancers
4. abnormal blood loss (maybe GI bleed)
5. pregnancy
6. ….

The problems is that these conditions (aside from being associated with anemia) are significant surgical risk factors- especially when coupled with other VERY common conditions- diabetes and hypertension. So the question: Is it the low HCT that increases mortality, or the presence of comorbidities??? I suspect that it is the diabetes, kidney disease, cancer etc that kills the patients, not the associated anemia…

So if the relationship between HCT and surgical death is CAUSATIVE, then the prevalence of comorbidities within the 3 groups (low HCT, nml HCT, high HCT) should be indistinguishable. If this were the case, preoperative management would reverse the trend, (i.e. make the anemic patients better, and normalize their risk…) This just doesn’t seem too plausible… The cancer-ridden diabetic still has cancer- it’s just that he has a few more RBC’s.

EDIT: Maybe there would still be differences in the prevalence of comorbidities even if the relationship was causative- but the risk of death should still fall out as independent from the presence of said medical condition..

In contrast, the if the relationship is correlative, then you might expect the low HCT high-risk patients to be sicker before surgery.. Under this scenario- you’d expect that it is not the anemia that is killing patients, but instead the presence of conditions, x, y, and z.. Here, treating the anemia does not make a difference in outcomes…

It seems reasonable to expect that the alcoholic cancer patient will have risk similar to other alcoholic cancer patients, regardless of whether they were transfused or not.. This seems especially likely given the relationship between VO2, DO2, and O2ER.

These data must exist… Somebody please! Does pre-operative management of anemia reduce risk of death form surgical complication?