Contra Hanson On Medical Effectiveness

🌈 Abstract

The article presents a detailed critique of Robin Hanson's argument that modern medicine does not work. It examines three major studies cited by Hanson - the RAND Health Insurance Experiment, the Oregon Health Insurance Experiment, and the Karnataka Health Insurance Experiment - and argues that they do not conclusively prove that medicine is ineffective. The article also presents additional evidence, such as improved survival rates for various diseases and other studies, that suggest medicine does in fact have significant positive effects.

🙋 Q&A

[01] Modern Medicine Improves Survival Rate

1. What evidence does the article provide to show that modern medicine improves survival rates?

• The article presents data showing significant improvements in 5-year survival rates for various types of cancer, as well as reductions in 30-day mortality rates for heart attacks and strokes, over the past few decades. These improvements are attributed to advances in medical treatments and interventions, rather than just changes in nutrition, sanitation, or wealth.

2. How does the article address potential counterarguments about age effects?

• The article acknowledges that improvements in survival could potentially be explained by people getting cancer/heart attacks at younger ages. However, it provides additional data showing similar survival improvements for childhood cancers and age-adjusted stroke mortality, indicating that age effects are not the primary driver of the observed survival gains.

[02] RAND Health Insurance Experiment

1. What were the key findings of the RAND Health Insurance Experiment?

• The RAND study found no significant effects of better insurance coverage on questionnaire-based health outcomes, smoking, weight, or cholesterol. However, it did find modest positive effects on vision and blood pressure.
• The article argues that the lack of effects on smoking, weight, and cholesterol are not surprising given the limited medical treatments available at the time. It also provides a more nuanced interpretation of the blood pressure results, suggesting they represent a real effect concentrated among those with hypertension.

2. How does the article assess the overall significance of the RAND study's findings?

• The article concludes that the RAND study does not prove that "medicine doesn't work", as Hanson claims. Rather, the authors argue the study's limitations in detecting health effects, especially for chronic conditions, make it an insufficient basis for such a strong conclusion.

[03] Oregon Health Insurance Experiment

1. What were the key findings of the Oregon Health Insurance Experiment?

• The Oregon study found that the group with Medicaid coverage used more medical care, had less medical debt, and reported better self-assessed health and mental health compared to the control group.
• However, the study did not find significant differences between the groups on objective health measures like blood pressure, cholesterol, and HbA1c.

2. How does the article interpret the lack of effects on objective health measures?

• The article argues that the lack of significant effects on objective measures can be explained by the study's limited power to detect changes that are consistent with the observed increases in medication use. It provides examples showing how plausible medication effects could have been missed due to the study's statistical limitations.

[04] Karnataka Health Insurance Experiment

1. What were the key findings of the Karnataka Health Insurance Experiment?

• The Karnataka study found that after adjusting for multiple comparisons, only 3 out of 82 health outcome measures were statistically significant. The authors could not reject the hypothesis that the distribution of p-values was consistent with no real differences between the groups.
• The article notes that the study found difficulties in the actual utilization of the insurance, with many participants unaware of how to use it or facing issues with hospitals and insurance companies.

2. How does the article interpret the Karnataka study's findings?

• The article argues that the lack of effects in the Karnataka study is likely due to the limited increase in actual healthcare utilization, rather than a true lack of medical effectiveness. It suggests the study's power was too low to rule out medically significant effects.

[05] Other, More Positive Studies

1. What additional evidence does the article provide showing positive effects of medicine?

• The article discusses a large-scale study on the effects of the Affordable Care Act's individual mandate, which found that sending a letter to uninsured individuals resulted in a lower mortality rate among the 45-64 age group.
• It also mentions other quasi-experimental studies that found reductions in mortality associated with Medicaid expansions and the implementation of Romneycare in Massachusetts.

2. How does the article respond to Hanson's dismissal of these positive studies?

• The article argues that Hanson's willingness to dismiss the positive findings from these larger, more powerful studies, while relying on the insurance experiments that are underpowered to detect meaningful effects, is inconsistent and unjustified.

[06] Final Thoughts

1. What is the article's overall assessment of the insurance experiment literature?

• The article acknowledges that the insurance literature does not provide a clear, unambiguous answer on whether extra health insurance has detectable population-level health effects. However, it argues that the evidence leans towards showing positive effects, especially when considering the larger and more robust studies.

2. How does the article respond to Hanson's claim that the insurance studies show medicine doesn't work?

• The article strongly rejects Hanson's claim, arguing that the insurance studies are simply too far removed from directly testing the effectiveness of medical treatments to support such a broad conclusion. It suggests the insurance studies are at best inconclusive about medication effectiveness, and should not be used to dismiss the substantial direct evidence that medicine does in fact have significant positive impacts.