Author + information
- Department of Cardiology, Ochsner Clinical School, University of Queensland, Australia; and the Ochsner Medical Center, New Orleans, Louisiana
- ↵∗Address for correspondence:
Dr. Christopher J. White, Department of Cardiology, Ochsner Medical Center, Ochsner Medical Center, 1514 Jefferson Highway, New Orleans, Louisiana 70121.
While attempting to create a pithy commentary on the scourge of peripheral arterial disease (PAD), I am struck by the seemingly slow progress we have made toward consistently applying therapies proven to lower the morbidity and mortality associated with this disease. The statistics are daunting. An estimated 8.5 million (approximately 1 in 20) U.S. adults have PAD, and the associated increased risks of adverse cardiovascular and limb-related outcomes result in a dramatic excess morbidity and mortality (1,2).
In this issue of the Journal, Hess et al. (3) present a well-written and informative analysis of outcomes in the year following a revascularization procedure for 381,415 patients with PAD between 2009 and 2015. Using the Premier Healthcare administrative database, which provides hospital-reported inpatient and hospital-based outpatient all-payer data from geographically diverse hospitals, they overcame many of the limitations of Medicare and inpatient-only datasets (4). Unique to this analysis is data on a broad PAD cohort, describing the outcomes for inpatient hospitalizations and outpatient endovascular peripheral revascularizations in the year following a procedure, and subsequently breaking down these hospitalizations into cardiovascular-related, limb-related, and others. Limb-related hospitalizations were further separated into major adverse limb events (MALE) (e.g., acute limb ischemia [ALI], including thrombectomy and thrombolysis; major amputation [at or above the ankle]; or surgical peripheral revascularization), inpatient endovascular revascularization, and other limb-related hospitalizations.
The authors’ observations reinforce the relatively grim 1-year outcomes that follow revascularization procedures for patients with symptomatic PAD (atypical and typical intermittent claudication in 70.7%, critical limb ischemia [CLI] in 26.2%, and ALI in 3.2%). One in 10 PAD limb revascularization patients was hospitalized within 1 year for a MALE. An additional 11% underwent an outpatient endovascular revascularization within that year, and 40% of patients had an all-cause inpatient hospitalization, of which one-half were limb-related and one-third were cardiovascular. Factors associated with increased risk for 1-year hospitalization for MALE included demographic factors (black, Medicaid and Medicare insurance), comorbid conditions (prior stroke, diabetes, renal insufficiency, heart failure, and smoking), and procedural factors (an index revascularization for CLI or ALI, a surgical or hybrid procedure, a procedure by a surgeon or radiologist, or discharged on oral anticoagulation).
The association of higher 1-year MALE events with treatment by a noncardiologist and a surgical or hybrid procedure can be partially explained by the higher acuity inpatient admissions for ALI or CLI likely more to be managed by surgeons or radiologists. However, variation in prescribing post-procedure antithrombotic therapy may have contributed to a higher post-procedure event rate. Fee-for-service Medicare patients, treated by surgeons or radiologists, are much less likely to receive a P2Y12 inhibitor after peripheral vascular intervention than if they were treated by cardiologists, and significant variation in the duration of prescribed therapy has been noted (5). The authors’ call for establishing multidisciplinary vascular teams, analogous to the “heart team” approach to transcatheter valve therapies, would be a positive step toward improving the overall care of patients with PAD.
This observational data exposes 1 particularly large elephant in the room: the clinicians’ failure to mitigate the adverse consequences of atherosclerotic PAD despite robust evidence supporting the benefits of guideline-directed medical therapies (GDMT) for PAD. For example, it is well established that statins prevent adverse secondary outcomes in atherosclerotic cardiovascular disease. For PAD specifically, the NHANES (National Health and Nutrition Examination Survey) reported statins to be associated with a 65% reduction in all-cause mortality (6).
Despite this evidence, the current analysis exposes significant underuse of GDMT in a post-revascularization population of patients with PAD (3). Only 61.7% of patients were discharged on statin therapy, 67.3% of patients were discharged on aspirin, 57.7% on a P2Y12 inhibitor, and 10.8% on a nonstatin lipid-lowering therapy. Although this underuse of risk-reduction strategies in PAD has been well documented, the reasons for the lack of adherence are complex (7). One major factor affecting physicians’ abilities to consistently perform tasks is the variation inherent in experts’ decision-making, initially described in the late 1960s by the psychologist Lewis R. Goldberg (8). In his groundbreaking studies of decision-making, Goldberg (8) found that “models of man” (linear algorithms) consistently outperformed “man” in reliably performing tasks. He attributed the experts’ inconsistency, when given the same situational stimulus, to a myriad of reasons—including boredom, fatigue, illness, and situational or interpersonal distractions. The algorithms, “models of man,” which were created from information provided by the experts, do not include this variation, with consistent results that outperform the experts.
Physicians’ variability in prescribing GDMT in patients with PAD was well documented in a recent report from the Veterans Affairs health care system, with only 41% of all patients with PAD (only 28.9% in patients with PAD without concomitant coronary disease) prescribed guideline-recommended statin therapy. This variability was accompanied by significant (20% to 28%) site level variation that was unrelated to specific patient factors (9).
There is evidence that algorithms created by experts are effective in improving clinical outcomes. In 1 study of hospitalized acute coronary syndrome (ACS) patients, an algorithm driven by computerized physician order entry with decision support (CPOE-DS), was compared with a group managed with a conventional ACS order set. The study measured compliance with guideline-based ACS performance measures (10). Patients managed with CPOE-DS were 5.7 times more likely to achieve all of the performance measures (i.e., perfect care [PC]) than those whose treatment was guided by the standardized order set. In this ACS population, achieving PC resulted in a significant survival advantage. The authors concluded that the use of CPOE-DS in a routine clinical setting was not only feasible, but very effective in attaining compliance with all of the performance measures (i.e., PC).
Given our lack of progress in reliably applying well accepted and evidence-based risk management strategies to reduce the morbidity and mortality of PAD, the time has come to act. To improve physicians’ adherence in following guidelines, Turley et al. (11) have called for stronger encouragement from professional societies and payers. But, as repeated surveys have shown, this strategy has been relatively ineffective in moving the field forward, with relatively low adoption rates of GDMT and significant individual and site variation (11). This should not be surprising. As Redelmeier and Tversky (12) have shown, physicians can know the right thing to do for a target group, but often fail to apply those standards to specific individuals.
Another alternative to improve physician compliance is the use of electronic health record–driven “alerts” or “pop-ups,” nudging clinicians toward optimal choices. In my experience, however, alert fatigue is a real factor preventing effective usage of a seemingly useful tool.
One promising strategy is educating patients and physicians on evidence-based practices. A report from a Canadian administrative PAD database demonstrated that an educational interventional program targeting patients, primary care physicians, and vascular surgeons with 8 major guideline-directed risk-management strategies dramatically improved cardiovascular and limb outcomes (13). Another educational technique combined transparent peer comparison with the use of accountable justification to effectively lower rates of inappropriate antibiotic prescribing (14). This technique would appear to be effective in increasing compliance with GDMT in a PAD population, but remains to be tested.
A final option to improve consistent application of GDMT for patients with PAD is for physicians to allow electronic health record–embedded order sets (algorithms) to drive standardized care, with the physician having the ability to edit the “pushed” orders to individualize patient care when appropriate.
Will physicians ever accept the substitution of clinical algorithms (cookbook medicine) for their individual clinical judgment? I think it is time to acknowledge our individual human fallibility largely due to the intrinsic unreliability of humans to provide consistent patient care without some form of automated support. This is not to say that clinician oversight of standardized care plans is not a key component of good patient care. There will always be some patients who will benefit from an individualized care plan. But for routine tasks, such as providing highly reliable, standardized, guideline-based patient care to patients with PAD, the evidence shows it is better left to the machines.
↵∗ 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.
Dr. White has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Benjamin E.J.,
- Virani S.S.,
- Callaway C.W.,
- et al.
- Hess C.N.,
- Rogers R.K.,
- Wang T.Y.,
- et al.
- ↵Premier Applied Sciences, Premier Inc. Premier Applied Sciences supports health care transformation through generation of real-world evidence. Available at: https://www.premierinc.com/transforming-healthcare/healthcare-performance-improvement/premier-applied-sciences/. Accessed May 28, 2018.
- Pande R.L.,
- Perlstein T.S.,
- Beckman J.A.,
- et al.
- Berger J.S.,
- Ladapo J.A.
- McBride C.L.,
- Akeroyd J.M.,
- Ramsey D.J.,
- et al.
- Turley R.S.,
- Mi X.,
- Qualls L.G.,
- et al.
- Hussain M.A.,
- Al-Omran M.,
- Mamdani M.,
- et al.