Technological development and medical productivity: the diffusion of angioplasty in New York state☆
Introduction
The dramatic technological change in the medical care industry has raised two key questions for patients, firms, and policymakers. First, what are implications of technological growth for the cost of care? Second, what are its effects on the quality of care? Considered in tandem, the answers to these questions illustrate the impact of technology on medical productivity. That is, they allow us to consider whether medical innovation is “worth it” in terms of increasing the level of quality-adjusted output per-unit of cost.
Much of the health economics literature subscribes to the view that technological innovation is a primary driver of increasing medical costs in the United States.1 Fuchs (1986) suggests that the American medical establishment is infused with a “technological imperative”—the belief that if a clinical intervention is possible and safe, it should be undertaken without regard for its costs and benefits. Other economic studies (Weisbrod, 1991, Newhouse, 1992, Cutler and McClellan, 1998) have also noted a positive relationship between technological growth and medical costs. What is somewhat puzzling about this positive relationship is that it appears to apply to many innovations that are actually less costly on a per-unit basis than existing technologies for which they are a substitute.
A key to understanding this puzzle is the phenomenon of treatment expansion—the provision of more intensive treatment to patients with low-grade symptoms (Cutler and McClellan, 2001). For example, prior to the early 1980s, patients with severe coronary artery disease (CAD) would receive coronary artery bypass graft (CABG) surgery, a major procedure that involves grafting a portion of vein or artery to bypass blockage in the coronary artery. Those with more mild CAD would undergo less costly (and less traumatic) medical management. The introduction of percutaneous transluminal coronary angioplasty (PTCA) in the late 1970s and early 1980s provided an intermediate treatment with cost and intensity levels that fell in between those of CABG and medical management. In contrast to CABG, PTCA involves only a small incision through which a balloon-tipped catheter is threaded. Upon reaching the point of blockage, the balloon is inflated to restore blood flow.
In this case, treatment expansion occurred among those PTCA patients who would have otherwise received only medical treatment.2 To the extent that some patients receive more intensive interventions, such expansion increases total costs and may have either positive or negative effects on outcomes. Treatment substitution—which allows a patient to shift from more- to less-intensive interventions—serves to offset the cost increases associated with expansion. Returning to the PTCA example, substitution involved some patients receiving PTCA in lieu of CABG. In theory, the effect of substitution—and the combined effect of expansion and substitution—on both costs and outcomes is ambiguous.
Beyond the ambiguity surrounding the offsetting effects of treatment expansion and substitution at a given point in time, one must also consider whether the relative magnitudes of these effects change over time. The equilibrium costs and benefits associated with many innovations are not evident at the time of their introduction (Gelijns and Rosenberg, 1994). As uncertainty is resolved over time—either through learning by doing or exogenous factors—the relationship between the marginal benefit and marginal cost of a technology may change. This may be the case for surgical procedures for which outcomes improve (and unit costs decline) with volume3 or for pharmaceuticals that are found to be effective for new indications over time. As such, the instantaneous productivity of a medical innovation measured shortly after its introduction might differ substantially from that measured at a later date.
In this paper, we address this issue of time-varying effects by examining the use of PTCA and CABG over the past two decades. PTCA is an innovation that is characterized by the uncertainty described above. Specifically, despite its advantages in terms of cost and patient comfort, PTCA initially was considered less effective than CABG for patients with relatively severe CAD. Due to learning and technological improvements over time, however, PTCA began to be used on increasingly severe cases that were previously reserved for CABG. In short, PTCA appears to have become a stronger substitute for CABG over time (Bohmer et al., 2000).
We use patient-level data from the State of New York to examine the utilization patterns for PTCA and CABG from 1982 to 2000. The beginning of this period roughly corresponds to the point at which PTCA was introduced into mainstream medical practice in the United States. We identify two distinct phases in the diffusion of PTCA. The 1980s and early 1990s constitute a period of treatment expansion during which the overall utilization rates for both PTCA and CABG increased. During the later 1990s, however, we find evidence of treatment substitution as the utilization rates for PTCA and CABG increased and decreased, respectively. While the introduction of PTCA raised total costs via treatment expansion, this increase was offset over time due to the substitution of PTCA for CABG. Though we must make several assumptions to consider the impact of PTCA on outcomes during the initial years of its diffusion, we find evidence that the expanded use of PTCA on patients who previously would have received only medical treatment appears to have improved the quality of care. Further, we present preliminary evidence suggesting that the substitution of PTCA for CABG eventually improved quality for patients receiving revascularization (i.e. either CABG or PTCA). Our results suggest that new technologies have the potential to become increasingly less costly over time and may result in productivity improvements.
The remainder of this paper is organized in four sections. Section 2 provides background on PTCA and its diffusion over the past two decades. Section 3 discusses issues related to our empirical methodology. Section 4 provides detail and descriptive analysis concerning the data from New York. Section 5 presents our results, and Section 6 concludes.
Section snippets
The development and growth of PTCA
Prior to the early 1980s, patients with CAD were faced with two possible paths of treatment. The decision between these paths typically depended on the severity of a patient’s illness. Those with relatively mild CAD might pursue medical management (i.e. they would not receive any invasive procedures), while those with more advanced disease would undergo CABG. As noted above, CABG was traumatic for patients and not infrequently resulted in death. Nonetheless, it was found to be a relatively
The impact of PTCA on cost
The first part of our empirical analysis considers the impact of PTCA’s diffusion on the cost of treating patients with CAD. We assume that the primary driver of PTCA’s impact on total cost is its effect on the rate at which both PTCA and CABG are provided for a fixed population. Increased PTCA use may also affect the cost-per-PTCA or cost-per-CABG due, for example, to volume–cost effects. While our analysis does not capture this latter impact on cost, it is likely that such volume–cost effects
Data and descriptive analysis
The main source of data for the cost analysis is New York’s Statewide Planning and Resource Cooperative System (SPARCS). SPARCS provides patient-level demographic (e.g. age, sex, payer type) and administrative (e.g. diagnoses, procedures performed, and total charges) information for all inpatient hospital discharges. The earliest year for which this data is available is 1982. Our analysis covers all relevant discharges in New York from 1982 to 2000 with the exception of 1983 and 1984.
Changes in utilization rates
The summary evidence presented thus far has certain limitations. In addition to facing potential selection problems, the CSRS and CARS data in Table 1 cover only a short period of time. While Fig. 2 does consider changes in CABG and PTCA rates per person, it does not provide evidence as to whether changes in CABG utilization are driven by changes in PTCA use or by some other characteristic of a given area. To address both of these issues, we take advantage of the information on county of
Discussion
Our analysis yields several conclusions. With respect to total costs, we find that an assessment of PTCA’s effects based solely on the first decade of its use yields a substantially different result than one—such as ours—that considers the entire life of the technology to date. Specifically, the growth of PTCA in the 1980s occurred predominantly through treatment expansion and was accompanied by little offset in the use of CABG and thus large increases in the overall cost of care. By the
Acknowledgements
We thank Thomas McGuire, Joseph Newhouse, two anonymous referees, and participants in the BU/Harvard/MIT Health Economics Seminar for helpful comments. We also thank Jeff Geppert for assistance in obtaining data.
References (42)
- et al.
High-tech capital formation and economic performance in US manufacturing industries: an exploratory analysis
Journal of Econometrics
(1995) - et al.
Randomised comparison of implantation of heparin-coated stents with balloon angioplasty in selected patients with coronary artery disease
Lancet
(1998) - et al.
The cost of angina pectoris in Sweden
PharmacoEconomics
(1995) - et al.
Will disruptive innovations cure health care?
Harvard Business Review
(2000) - Bypass Angioplasty Revascularization Investigation (BARI) Investigators, 1996. Comparison of coronary bypass surgery...
- CABRI Trial Participants, 1995. First-year results of CABRI (Coronary Angioplasty Versus Bypass Revascularization...
- Center for Evaluative and Clinical Sciences, 1999. Dartmouth atlas of health care in the United States....
- et al.
Myocardial infarction and cardiac mortality in the Bypass Angioplasty Revascularization Investigation (BARI) randomized trial
Circulation
(1997) - et al.
Cost-effectiveness of coronary stenting in acute myocardial infarction: results from the stent primary angioplasty in myocardial infarction (Stent-PAMI) trial
Circulation
(2001) - Cutler, D., McClellan, M., 1998. What is technological change? In: Wise, D. (Ed.), Inquiries in the Economics of Aging....
Is technological change in medicine worth it?
Health Affairs
Measuring the health of the US population
Brookings Papers on Economic Activity Microeconomics
ACC/AHA guidelines for coronary artery bypass graft surgery: executive summary and recommendations
Circulation
Substituting diagnostic services: new tests only partly replace older ones
Journal of the American Medical Association
Volume-outcome relationships and inhospital mortality: the effect of changes in volume over time
Medical Care
Can technology assessment control health spending?
Health Affairs
The dynamics of technological change in medicine
Health Affairs
Variation in the use of cardiac procedures after acute myocardial infarction
New England Journal of Medicine
Coronary artery bypass graft surgery: the relationship between inhospital mortality rate and surgical volume after controlling for clinical risk factors
Medical Care
Coronary angioplasty volume-outcome relationships for hospitals and cardiologists
Journal of American Medical Association
Cited by (0)
- ☆
The views presented in this paper do not necessarily reflect the findings of the New York State Department of Health or the New York State Cardiac Advisory Committee.