The NBER working paper Technology Growth and Expenditure Growth in Health Care, by Amitabh Chandra and Jonathan S. Skinner, includes, among other things, a helpful typology of health care technology. While it may be correct to say that health care cost inflation is related to health care technology, it is important to recognize that not all health care technology is the same. Chandra and Skinner organize technology into three types:
Type I: “Home Run” Technologies. “The first category of medical technologies includes those that are cost-effective and useful for nearly everyone in the relevant population.” These include*
- The washing of hands before surgery,
- The use of gloves,
- Cleansing wounds,
- Surfactants to treat neonatal acute respiratory distress for newborns,
- Antiretroviral drugs for HIV patients,
- Testicular cancer treatment.
The authors point out that not all type I treatments are low cost. However, the high cost ones are cost-effective, dramatically enhancing and extending life for those requiring them. Moreover, few would choose such treatments if they were not necessary. Amusingly, Chandra and Skinner write, “[F]ew men would agree to undergo an orchiectomy (don’t ask) without a confirmed diagnosis, limiting the growth in expenditures on testicular cancer.” Quite right!
Type II: Potentially Cost-Effective Technologies with Heterogeneous Benefits. These are treatments that “are cost-effective in some patients but have declining marginal benefits in others. […] Despite clear benefits to some patients, such technologies can still exhibit modest or even poor average cost effectiveness across all patients.” These include*
- Angioplasty (improve heart attack survival only if performed within a day of onset),
- Cesarean sections,
- Imaging technologies,
- Prostate cancer screening and treatment.
Unlike type I treatments, those of type II are often applied to patients for whom they do not provide benefit. That is, sometimes type II treatments are unnecessary care. As such, the overall cost-benefit of treatments in this category is lower than for type I treatments. That is not to say such treatments are not beneficial. They are, but only when applied appropriately, and they often are not.
Type III: Technologies with Modest or Uncertain Effectiveness. These are treatments for which “the average value of the procedure leads to poor (or non-existent) cost-effectiveness, or where there is considerable uncertainty about its benefits.” They include*
- Arthroscopic surgery for osteoarthritis of the knee,
- Frequent office visits,
- ICU days for the chronically ill,
- Referrals to specialists.
Informed by this typology, and other considerations I have not written about above, the paper concludes,
[A]ttributing cost growth and improvements in outcomes to “technology growth” is too simplistic and tells us little about where the cost growth is occurring, whether such growth should be tamed, and if so, how it should be done. […]
[T]here is wide heterogeneity in the productivity of medical treatments, ranging from very high (aspirin for heart attacks and surfactants for premature births) to low (stents for stable angina), or simply zero (arthroscopy for osteoarthritis of the knee). […] It is perhaps tautological, but still worth noting that countries or systems of care that encourage the first group of innovations but discourage the third group are most likely to exhibit high aggregate productivity growth and a slower overall [cost] growth rate relative to GDP.
Finally, the most difficult questions relate to health care policy going forward, where there does not appear to be a single magic bullet to “solve” the health care problem. The U.S. and other countries are struggling with rising costs and a diminished ability to raise taxes or health insurance premiums. As well, the U.S. leads the world with the extent of waste in their health care system (McKinsey Global Institute, 2008). The extent of waste in the U.S. could, ironically, prove to be a boon if a fundamental restructuring of health care unleashed some of this lost productivity; gradually eradicating 30 percent of waste would depress cost growth rates by 1.3 percentage points over the next two decades. The alternative to not making such changes is far more worrisome: Rising political and economic resistance against tax hikes, insurance premium increases, or coverage expansion could serve as particularly inefficient brakes on both health care costs and health care innovation.
* These are the authors’ examples. Perhaps some categorizations are debatable.