Author + information
- James G. Jollis, MD, FACC⁎ ()
- ↵⁎Reprint requests and correspondence to:
Dr. James G. Jollis, Box 3254, Duke University Medical Center, Durham, North Carolina 27710
Nam in omnibus fere minus ualent praecepta quam experimenta. [In almost everything, experience is more valuable than precept.]
—Marcus Fabius Quintilian, De Institutione Oratoria (II, 5, 15) (1)
An 80-year-old woman is found by paramedics with crushing chest pain, a systolic blood pressure of 60 mm Hg in the left arm and 80 mm Hg in the right arm, and 4 to 6 mm of ST-segment elevation across the precordial leads. The medics have 2 options: transport to Hospital A, 5 min away, which routinely performs diagnostic coronary angiography and occasionally performs coronary angioplasty in urgent situations, or transport the patient to Hospital B, 30 min away, which performs more than 600 angioplasties per year. In concert with local cardiologists, the emergency medical system director has established a pre-specified “destination plan” directing the medics to transport patients with ST-segment elevation myocardial infarction (STEMI) to one of these hospitals.
Since the original observation by Hal Luft in 1978 showing fewer deaths among patients undergoing procedures at greater-volume hospitals, the issue of “volume and outcome” has vexed the medical profession (2). In 1988, the American College of Cardiology first adopted a physician volume standard of “about 1 case/week” to maintain proficiency (3). This seemingly minimal standard unleashed a firestorm of controversy by physicians concerned about exclusion from angioplasty based upon a guideline lacking empirical evidence. In response to the call for supporting evidence, the standards were rewritten 5 years later after careful review of volume-outcome relationships in available angioplasty registries and claims data. According to an inflection point for increasing mortality below 75 cases/year, the guidelines increased the minimum operator volume standard and recommended at least 200 cases/year for hospitals (4). The reaction of some interventionalists remained forceful, as exemplified by the appearance of buttons with a red slash through number 75 at national meetings. Most recently, with the emerging national mandate to provide prompt primary angioplasty to all patients with STEMI, the benefit of transferring patients to high-volume regional percutaneous coronary intervention (PCI) centers must be weighed against the additional risk of delayed treatment. The volume-outcome relationship is most apparent among high-risk patients, and STEMI patients have the most to gain from experienced PCI teams and the most to lose from delays (5,6).
In this issue of the Journal, Srinivas et al. (7) provide an analysis of the volume-outcome relationship for primary PCI in contemporary data. This study is remarkable for a number of reasons. The New York State registry likely represents the best available data by which to examine volume and outcome in the U.S. Unlike national registries, participation in the New York State PCI registry is mandatory and global, so that all operators and hospitals, including lower-volume institutions, must participate (8). Thus, the New York data offer a greater opportunity to consider PCI in widespread practice. Other significant advantages of the New York data compared with national registries involve data verification. Rather than relying on self-reported complications and mortality, the New York registry is cross-referenced with state vital status data, resulting in verified and highly reliable mortality measures. Patient risk factors also are subject to verification by chart review by the Department of Health.
The work, in which 7,321 patients undergoing primary PCI are examined, identifies significant associations between procedural volume and in-hospital mortality according to physician and hospital volume. The New York data also identify interactions between physician and hospital volume, with trends for better outcomes for patients treated by greater-volume physicians at greater-volume institutions. The finding of an interaction between physician and hospital volume is not uniform, but small “cell size” issues, particularly in stratifying patients treated by high-volume physicians in low-volume hospitals, likely contribute to heterogeneity in these observations. Thus, 25 years after Luft's original observation, some of the best data available in coronary revascularization continue to identify mortality differences according to physician and hospital volume for primary angioplasty.
What is to be done with these data? First, the data must be viewed in the context of an observational study. Patients treated at lower-volume hospitals and by lower-volume physicians were sicker according to a number of measures. No amount of adjustment in regression models can completely separate the greater illness severity from worse outcomes. This finding suggests that some portion of the relationship may be due to selection bias. Outcomes may appear better at high-volume hospitals if the sickest patients are deemed too unstable to transfer or die in transfer before reaching their catheterization laboratories.
Similarly, the relationship between physician volume and hospital volume is highly confounded, and the relative contribution of each to better outcomes cannot be completely defined. One of the most significant criticisms of the application of volume-outcome observations to public policy involves the potential exclusion of “quality” low-volume providers. Because of small numbers, statistical techniques are unable to differentiate low-volume providers with above-average skills from low-volume providers with above-average luck. Because patients “on average” fare better with greater-volume physicians and hospitals, guidelines and public policy that take into account volume will lead to the best overall outcomes.
The last important issue raised by the findings of Srinivas et al. (7) involves the underlying causes of the consistent finding of lower mortality for patients treated by greater-volume providers. In addition to differences in illness severity and selection bias noted previously, the most plausible explanation involves experience. In practical terms, the difference between survival and death often involves a PCI team that can recognize and manage severe complications. When coronary interventions are proceeding well, modern techniques and devices are capable of restoring coronary flow in an expedient and successful fashion. The value of experience does not become apparent until the interventional team is called upon to manage infrequent but extreme situations. Following the patient example mentioned previously, any number of situations related to inexperience could lead to a poor outcome: failure to recognize that a guide catheter has entered a false lumen and is extending an aortic dissection; inadequate resuscitative efforts with weak chest compressions and slow intra-aortic balloon pump preparation and deployment; or “back walling” the femoral artery in a hypotensive patient with diminished femoral pulses leading to a large retroperitoneal bleed.
Although experience, selection bias, and illness severity all contribute to the observed differences in mortality rates, these data continue to suggest that primary angioplasty is best performed on a regular basis by experienced cardiac teams. Returning to the issues of STEMI systems raised by our patient, most primary angioplasty facilities can be staffed and prepared to treat patients within 30 min. Thus, policies should generally direct STEMI patients to greater-volume facilities that are within 30 min if these facilities are routinely capable of rapid activation by referring emergency physicians and paramedics on a 24 h/day, 7 days/week basis. To meet the needs of our patients and fulfill the public policy mandate of these consistent observations, experienced physicians and hospitals should make every effort to provide primary PCI on a timely and round-the-clock basis.
Dr. Jollis has received grant support from United Healthcare, Genentech, and Sanofi-Aventis.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Reinhardt T.,
- Winterbottom M.
- Ryan T.J.,
- Faxon D.P.,
- Gunnar R.M.,
- et al.
- Ryan T.J.,
- Bauman W.B.,
- Kennedy J.W.,
- et al.
- Jollis J.G.,
- Peterson E.D.,
- Nelson C.L.,
- et al.
- Hannan E.L.,
- Wu C.,
- Walford G.,
- et al.
- Srinivas V.S.,
- Hailpern S.M.,
- Koss E.,
- Monrad E.S.,
- Alderman M.H.
- Nallamothu B.K.,
- Wang Y.,
- Curtis J.P.,
- et al.