HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Robert D. Stewart Sidney Levitsky Stephen J. Lahey Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stewart, R. D. Articles by Lahey, S. J. Search for Related Content PubMed PubMed Citation Articles by Stewart, R. D. Articles by Lahey, S. J.

Ann Thorac Surg 2000;70:169-174
© 2000 The Society of Thoracic Surgeons

---

Original articles: Cardiovascular

Predictors of 30-day hospital readmission after coronary artery bypass

^a Division of Cardiothoracic Surgery, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA

Address reprint requests to Dr Lahey, Division of Cardiothoracic Surgery, Beth Israel Deaconess Medical Center, Suite 2C, 110 Francis St, Boston, MA 02215
e-mail: sl4cabg{at}aol.com

	Abstract

Top Abstract Introduction Material and methods Results Comment References

 
Background. Risk factors for 30-day hospital readmission following coronary artery bypass grafting (CABG) have not been established.

Methods. We prospectively followed 485 consecutive patients who underwent isolated primary CABG at our institution in 1997. Patients were contacted by telephone at 30 days following operation to determine readmission status.

Results. The overall readmission rate was 16% (76 of 485). Female gender (25% versus 11%, p = 0.001) and diabetes (22% versus 12%, p = 0.005) were associated with significantly higher readmission rates. The relationship between female gender and readmission persisted after correcting for age and other comorbidities. Congestive heart failure trended towards a significant relationship with increased readmission rate (22% versus 14%, p = 0.09). There were no significant associations between 30-day readmission rate and age, hypertension, chronic obstructive pulmonary disease, history of myocardial infarction, peripheral vascular disease, creatinine level of 1.4 mg/dL, or decreased left ventricular ejection fraction (< 40%).

Conclusions. These data show that most of the classic risk factors for postoperative mortality are not necessarily associated with increased readmission. However, female gender and diabetes are associated with greater than twice the risk of 30-day readmission following CABG.

	Introduction

Top Abstract Introduction Material and methods Results Comment References

 
The emergence of health maintenance organizations (HMO) and changes in Medicare reimbursement schedules have prompted a revision of how the cost of a medical intervention is measured. In the 1990s, many hospitals were forced to negotiate preset, global fees for individual services under fully or partially capitated contractual arrangements with HMOs. Traditionally, strategies used by hospitals to reduce inpatient costs were directed toward reducing patient length of stay, while at the same time maximizing current procedural terminology (CPT) and diagnostic related group (DRG) coding. In the future, Medicare and/or HMO reimbursement may be linked to clinical outcome in such a way that the overall cost incurred by the hospital will include any hospital readmission by the patient within a predefined period of time (eg, 30 days). This new economic burden will become increasingly important for those procedures associated with high 30-day readmission rates, such as coronary artery bypass grafting (CABG).

Although little is known of true 30-day readmission rates following cardiac operation, recent studies have reported rates ranging from 7.1% to 20.9% [1–5]. Efforts to decrease the cost of providing complete health care for CABG patients must include reducing readmission rates following CABG and consequently, preventing de facto cost shifting to other providers. An understanding of which characteristics place patients at increased risk for readmission is essential but is, to date, unknown.

This study was undertaken to evaluate whether several established risk factors for increased morbidity and mortality following CABG are also associated with an increased risk of 30-day readmission. These variables include older age, female gender, prior myocardial infarction (MI), congestive heart failure (CHF), reduced ventricular function, chronic obstructive pulmonary disease (COPD), renal insufficiency, peripheral vascular disease (PVD), and diabetes (DM).

	Material and methods

Top Abstract Introduction Material and methods Results Comment References

 
This study included 503 consecutive patients who underwent isolated primary CABG between January 1 and December 31, 1997 at our institution. Patients who died while in the hospital ( ) and surviving patients who had not been discharged by postoperative day 30 ( ) were excluded from this analysis. The final cohort included 485 patients.

All patients were personally contacted by telephone by the authors on the 30th postoperative day. Patients were asked if they had been to any hospital since their discharge. If patients were unable to communicate this information, a family member or the patient’s personal physician was contacted to determine if the patient had been readmitted. Consent was obtained from each patient to acquire the relevant hospital records documenting any readmissions. All readmissions, as opposed to emergency room visits, were confirmed by review of discharge summary. Readmission diagnosis was determined from these records. All readmissions for "chest discomfort" were categorized as "angina" regardless whether or not the etiology was demonstrated to be ischemia related. There were no protocols for admission criteria. Each admission was at the discretion of the admitting cardiologist, internist, or emergency physician. Follow-up was obtained for all 485 patients.

Patient medical data were abstracted from hospital records. Data collected included age, gender, history of DM, history of MI, history of hypertension, history of CHF, history of COPD, history of PVD, base line serum creatinine, and left ventricular ejection fraction (LVEF). LVEF was known, by either echocardiography or by ventriculography, in 460 of the 485 patients. Postoperative records were also reviewed to determine if each patient experienced new onset atrial fibrillation following their CABG. Atrial fibrillation was defined as any documented occurrence that lasted more than 1 hour or that required therapeutic intervention regardless of duration. The patient data are shown in Table 1.

Statistical analysis was performed using STATA software (Stata, College Station, TX). Continuous variables were compared using Student’s t tests. Categorical or dichotomous variables were compared using Pearson ² or Fisher’s exact tests where appropriate. Multivariate logistic regression was performed with all variables included in the model to control for potential confounding or effect modification. Statistical significance was defined as a 2-sided p value of less than 0.05.

	Results

Top Abstract Introduction Material and methods Results Comment References

 
Seventy-six patients (16%) were readmitted to a hospital within 30 days of CABG and 409 patients (84%) were not readmitted (Fig 1). The readmission diagnoses are listed in Table 2. The most common reason for readmission was atrial fibrillation followed by leg wound infection. Congestive heart failure or angina was the third most common cause for readmission.

View larger version (14K): [in this window] [in a new window]   Fig 1. The 30-day readmission status.

View larger version (13K): [in this window] [in a new window]   Fig 2. Patient characteristics associated with increased 30-day readmission. (CHF = congestive heart failure; Hx = history.)

View larger version (18K): [in this window] [in a new window]   Fig 3. Patient characteristics not associated with 30-day readmission. (Cr = base line serum creatinine; COPD = chronic obstructive pulmonary disease; HTN = hypertension; LVEF = left ventricular ejection fraction; MI = myocardial infarction; PVD = peripheral vascular disease.)

Length of stay was inversely associated with the 30-day readmission. Among the 347 patients with length of stay less than 7 days there were 44 readmissions (13%) compared with 32 readmissions among the 138 patients with a length of stay 7 days or longer (23%, p = 0.004). Patients who were discharged alive by postoperative day number 4 had a lower readmission rate than patients with any other length of stay, only 7 of 78 patients (9%).

Women and diabetic patients were more likely to be readmitted for any reason by 30 days compared with men and nondiabetic patients; however, the most striking difference was in the rates of readmission in these two groups for wound infections. For women, 35% of readmissions were for wound infections (13% sternal, 22% leg), whereas only 15% of male readmissions were for wound infections (0% sternal, 15% leg). Similarly for diabetic patients, 34% of readmissions were for wound infections (10% sternal, 24% leg) compared with only 16% of nondiabetic readmissions (3% sternal, 13% leg). However, because of the relatively small number in each readmission diagnosis category these differences were not statistically significant. The study was powered to examine the overall readmission rate and not to determine differences in specific readmission diagnoses.

Multiple logistic regression (Table 3) controlling for all of the measured patient characteristics demonstrates that female gender remained a highly significant predictor of increased readmission rate with an odds ratio of 2.4. This odds ratio compares favorably with the raw data, which demonstrated that the women had greater than twice the rate of 30-day readmission as the men. Diabetes had borderline statistical significance after controlling for the other comorbidities in the regression analysis. None of the other preoperative factors were significantly associated with readmission rate.

Among patients who developed postoperative atrial fibrillation during the original hospitalization and were readmitted, the most common readmission diagnosis was atrial fibrillation (10 of 25, 40%). This finding was significantly higher than the incidence of atrial fibrillation among those readmitted who did not experience atrial fibrillation during initial hospitalization (7 of 51, 13%, p = 0.04).

	Comment

Top Abstract Introduction Material and methods Results Comment References

 
Gender has become an increasingly important issue in cardiac medicine and coronary operations in the past several years. Issues of treatment bias have arisen, and as a result, increasing numbers of women are being referred for and are undergoing CABG. Most studies on CABG outcome have also shown women to have greater rates of morbidity and mortality [6, 7]. Whether female gender, per se, or the relatively smaller body surface area of women is the true risk factor is not clear.

The most significant finding of this study is that women were readmitted more than twice as often as men within 30 days of CABG. Why is this the case?

The most simple hypothesis is that women may develop more complications following discharge. However, this jump in logic would suggest that all of the risk factors for increased complications, such as age more than 70 years, would also portend a higher readmission rate. This was clearly not the case in this study.

Another possible explanation for the observed differences in readmission rates is that women are readmitted more often than men because of some poorly defined gender bias. Perhaps physicians are more likely to readmit a woman presenting with a given complaint following CABG than a man presenting with the same complaint.

Finally, there may be a gender-based difference in behavior following CABG. Women may be more likely than men to present to their physician or to an emergency room with a given complaint and subsequently be readmitted.

The same possible hypotheses apply to diabetic patients. Perhaps diabetic patients have more complications. Perhaps there is a bias among physicians toward readmitting diabetic patients more than nondiabetic patients, reflecting physicians’ unwillingness to tolerate even minor complications in diabetic patients. Finally, it is possible that diabetic patients are simply more likely to seek medical attention.

The association of both female gender and DM with increased readmission found in this study was strong. The clinical difference was large; women and diabetic patients each had greater than twice the readmission rate of men and nondiabetic patients, respectively. The statistical association was strong as well. The study made multiple comparisons of the readmission rate and numerous patient variables, and therefore, a p value of 0.05 may be due to chance. But p values of 0.001 and 0.005 cannot be ascribed to such statistical wrangling. Furthermore, female gender retained a high level of significance after controlling for all other variables, including older age, in the multivariate logistic regression.

Unfortunately, definitive explanations of why women and diabetic patients have higher readmission rates cannot be gleaned from this study. The differences in the rates of readmission for various diagnoses, that is, increased wound infection rates in women and diabetic patients, is suggestive. However, this study was not powered to, and did not find, a statistically significant difference between genders or between diabetic and nondiabetic patients in terms of particular readmission diagnoses. These data warrant a larger study with possible interventions aimed at decreasing these problems. Such interventions could include various less invasive incisions (interrupted or endoscopic), better dressings (Ace wraps or compression stockings to reduce edema), or prophylactic antibiotics. This particular study does not shed light on these potential solutions but does highlight the problem that needs to be addressed.

An equally interesting finding in this study was the consistency of the 30-day readmission rates for most of the other characteristics that have been shown to be risk factors for adverse events following CABG. The presence of older age, decreased ventricular function, PVD, COPD, hypertension, prior MI, or renal insufficiency did not significantly increase patients’ risk of readmission. Some small clinical differences did exist for PVD and Cr levels of more than 1.4 mg/dL and might represent real differences in readmission rate; however, the relatively small number of patients with these comorbidities preclude any discrimination between chance and the existence of a real difference because of insufficient statistical power. Type II error aside, the clinical relevance of these factors is small, with odds ratios of approximately 1.3. Another unexpected result noted in this analysis was the apparent lower incidence of readmission in the patients with decreased LVEF. As the p value of 0.3 suggests, this was likely a spurious result.

The inverse relationship of length of stay and readmission rate demonstrates that we have been able to identify patients who can safely be discharged early. Conversely, slightly increased length of stay is at least a marker of higher risk for readmission. These data do not address whether increasing the length of stay for each individual, or decreasing the length of stay, would effect the readmission rate.

The strength of this study is the 100% capture and 100% follow-up of isolated primary CABG patients at a single institution over an entire year. All patients were accounted for and all readmissions were confirmed by hospital discharge summary. This accounting may not be necessary in other outcome research projects in which the data for 70% or 80% of the population is representative of those lost to follow-up. However, for 30-day readmission analysis, loss to follow-up is probably associated with the outcome itself, because readmitted patients are often the ones hardest to locate and thus are "lost to follow-up." This point is made more relevant by the fact that approximately 50% of the readmitted patients in this study were readmitted to outside hospitals.

That 50% of the readmissions were to outside institutions limits our ability, as the primary operating institution, to influence the decision to readmit. The discretion of the admitting physician may account for a possible bias in readmissions as suggested above, or may account for many nonindicated readmissions. It was not possible to determine the validity of each readmission, though some were questionable (eg, 1 patient was readmitted for wound pain).

The decision to include only isolated, primary CABG patients (and thus exclude redo operations and CABG procedures in association with valve, ascending aortic, or carotid operations) was made to eliminate any bias of these procedures on the outcome. Because of the small number of these cases relative to isolated, primary CABG cases, their inclusion would have added little statistical power to the study. The number of these cases was insufficient to perform subgroup analyses. Moreover, the increased complexity of these other procedures might have led to differences in readmission rates that could have biased the analysis of the isolated, primary CABG cases. The decision to include only those patients undergoing a single operation, together with the relatively large numbers of isolated, primary CABG cases, increases the validity of these findings and makes for robust statistical analysis. The limitation of this exclusion is that the results observed in this study may not be generalizable to cardiac procedures other than isolated, primary CABG.

Comprehensive health care costs are becoming increasingly difficult to determine as providers widen the scope of delivery systems to include outpatient ancillary services and other directly related care. A significant and persistent impediment to both the analysis of readmission trends and the control of health care expenditures may be the inherently opposing incentives of the outside hospital and those of the tertiary center where the original procedure was performed. In our study approximately 50% of readmitted patients were readmitted to outside hospitals. If Medicare reimbursement for CABG ultimately becomes covered by a global, preset fee, the financial incentive to readmit a postoperative CABG patient may disappear. In the future, the enormous cost of readmitting a patient after CABG may become as important to the hospitals providing the initial care as the initial length of stay is today.

Although gender and DM cannot be altered or eliminated to decrease the risk of 30-day readmission, a more focused clinical approach may be warranted in these patients to reduce their risk of readmission. A recent report from Sweden showed that 1-year readmission rates were 14% in a study population who were enrolled in a rehabilitation program as compared with a 32% readmission rate in the control group [8]. Whereas such a program may be beneficial to all patients, it may be particularly cost-effective in those patients with the greatest risk of readmission: female and diabetic patients.

The incidence of postoperative atrial fibrillation, unlike gender and DM, can potentially be altered. Atrial fibrillation is commonly recognized as the most frequent complication following CABG and leads to significantly increased hospital length of stay. Any possibility of substantially reducing the readmission rate associated with atrial fibrillation can be achieved only through a better means of prophylaxis or treatment and control of this common arrhythmia following CABG. This imperative is particularly significant given the high incidence of readmission for atrial fibrillation among patients who experienced atrial fibrillation during the initial hospitalization. Whether attempting to prevent the initial episode or aggressively treating the patients who experience atrial fibrillation to prevent recurrences is a question these data cannot answer.

Beyond intervention on the patients themselves, the decision processes that determine readmission must be examined. If women are being readmitted more often than men because of a gender bias, then an analysis and eventual understanding of this decision process might decrease readmissions that are not considered necessary in comparable male patients. Similarly, we must examine whether there exists a bias among physicians to "automatically" readmit diabetic patients with minor complications. If such a bias exists, it must be determined whether this lower readmission threshold is appropriate in the setting of DM.

	References

Top Abstract Introduction Material and methods Results Comment References

 

1. Engelman R.M., Rousou J.A., Flack J.E., et al. Fast track recovery of the coronary artery bypass patient. Ann Thorac Surg 1994;58:1742-1746.[Abstract/Free Full Text]
2. Hennen J., Krumholtz H.M., Radford M.J., Meehan T.P. Readmission rates, 30 days and 365 days postdischarge among the 20 most frequent DRG groups, Medicare inpatients age 65 or older in Connecticut Hospitals, fiscal years 1991, 1992, and 1993. Conn Med 1995;59:263-270.[Medline]
3. Lahey S.J., Campos C.T., Jennings B., Pawlow P., Stokes T., Levitsky S. Hospital readmission after cardiac surgery. Circulation 1998;98(Suppl):II35-II40.
4. Beggs V.L., Birkmeyer N.J., Nugent W.C., Dacey L.J., O’Connor G.T. Factors related to rehospitalization within thirty days of discharge after coronary artery bypass grafting. Best Pract Benchmarking Healthcare 1996;1:180-186.
5. Stanton B.A., Jenkins C.D., Goldstein R.L., Vander Salm T.J., Klein M.D., Aucoin R.A. Hospital readmissions among survivors six months after myocardial revascularization. JAMA 1985;253:3568-3573.[Abstract/Free Full Text]
6. O’Connor G.T., Morton J.R., Diehl M.J., et al. Differences between men and women in hospital mortality associated with coronary artery bypass graft surgery. Circulation 1993;88:2104-2110.[Abstract/Free Full Text]
7. Edwards F.H., Carey J.S., Grover F.L., Bero J.W., Hartz R.S. Impact of gender on coronary bypass operative mortality. Ann Thorac Surg 1998;66:125-131.[Abstract/Free Full Text]
8. Perk J., Hedback B., Engvall J. Effects of cardiac rehabilitation after coronary artery bypass grafting on readmissions, return to work, and physical fitness. Scand J Soc Med 1990;18:45-51.[Medline]

This article has been cited by other articles:

				M. J. Mack, A. P. Banning, P. W. Serruys, M.-C. Morice, Y. Taeymans, G. Van Nooten, G. Possati, F. Crea, K. L. Hood, K. Leadley, et al. Bypass Versus Drug-Eluting Stents at Three Years in SYNTAX Patients With Diabetes Mellitus or Metabolic Syndrome Ann. Thorac. Surg., December 1, 2011; 92(6): 2140 - 2146. [Abstract] [Full Text] [PDF]

				E. L. Hannan, Y. Zhong, S. J. Lahey, A. T. Culliford, J. P. Gold, C. R. Smith, R. S. D. Higgins, D. Jordan, and A. Wechsler 30-Day Readmissions After Coronary Artery Bypass Graft Surgery in New York State J. Am. Coll. Cardiol. Intv., May 1, 2011; 4(5): 569 - 576. [Abstract] [Full Text] [PDF]

				B. M. Murphy, P. C. Elliott, M. R. Le Grande, R. O. Higgins, C. S. Ernest, A. J. Goble, J. Tatoulis, and M. U. C. Worcester Living alone predicts 30-day hospital readmission after coronary artery bypass graft surgery European Journal of Cardiovascular Prevention & Rehabilitation, April 1, 2008; 15(2): 210 - 215. [Abstract] [Full Text] [PDF]

				C. R. Parikh, S. G. Coca, G. L. Smith, V. Vaccarino, and H. M. Krumholz Impact of Chronic Kidney Disease on Health-Related Quality-of-Life Improvement After Coronary Artery Bypass Surgery. Arch Intern Med, October 9, 2006; 166(18): 2014 - 2019. [Abstract] [Full Text] [PDF]

				V. Guru, S. E. Fremes, P. C. Austin, E. H. Blackstone, and J. V. Tu Gender Differences in Outcomes After Hospital Discharge From Coronary Artery Bypass Grafting Circulation, January 31, 2006; 113(4): 507 - 516. [Abstract] [Full Text] [PDF]

				K. Dickstein, G. W. Gleim, S. Snapinn, M. K. James, J. Kjekshus, and for the OPTIMAAL Study Group The impact of morbid events on survival following hospitalization for complicated myocardial infarction Eur J Heart Fail, January 1, 2006; 8(1): 74 - 80. [Abstract] [Full Text] [PDF]

				E. L. Hannan, C. Wu, G. Walford, S. B. King III, D. R. Holmes Jr, J. A. Ambrose, S. Sharma, S. Katz, L. T. Clark, and R. H. Jones Volume-Outcome Relationships for Percutaneous Coronary Interventions in the Stent Era Circulation, August 23, 2005; 112(8): 1171 - 1179. [Abstract] [Full Text] [PDF]

				A. Ergul, J. S. Johansen, C. Stromhaug, A. K. Harris, J. Hutchinson, A. Tawfik, A. Rahimi, E. Rhim, B. Wells, R. W. Caldwell, et al. Vascular Dysfunction of Venous Bypass Conduits Is Mediated by Reactive Oxygen Species in Diabetes: Role of Endothelin-1 J. Pharmacol. Exp. Ther., April 1, 2005; 313(1): 70 - 77. [Abstract] [Full Text] [PDF]

				O. Jarvinen, J. Julkunen, T. Saarinen, J. Laurikka, and M. R. Tarkka Effect of Diabetes on Outcome and Changes in Quality of Life After Coronary Artery Bypass Grafting Ann. Thorac. Surg., March 1, 2005; 79(3): 819 - 824. [Abstract] [Full Text] [PDF]

				R. A. DiMaria-Ghalili Changes in Body Mass Index and Late Postoperative Outcomes in Elderly Coronary Artery Bypass Grafting Patients: A Follow-up Study Biol Res Nurs, July 1, 2004; 6(1): 24 - 36. [Abstract] [PDF]

				V. Vaccarino, Z. Q. Lin, S. V. Kasl, J. A. Mattera, S. A. Roumanis, J. L. Abramson, and H. M. Krumholz Sex Differences in Health Status After Coronary Artery Bypass Surgery Circulation, November 25, 2003; 108(21): 2642 - 2647. [Abstract] [Full Text] [PDF]

				N. Luciani, G. Nasso, M. Gaudino, A. Abbate, F. Glieca, F. Alessandrini, F. Girola, F. Santarelli, and G. Possati Coronary artery bypass grafting in Type II diabetic patients: a comparison between insulin-dependent and non-insulin-dependent patients at short- and mid-term follow-up Ann. Thorac. Surg., October 1, 2003; 76(4): 1149 - 1154. [Abstract] [Full Text] [PDF]

				E. L. Hannan, M. J. Racz, G. Walford, T. J. Ryan, O. W. Isom, E. Bennett, and R. H. Jones Predictors of Readmission for Complications of Coronary Artery Bypass Graft Surgery JAMA, August 13, 2003; 290(6): 773 - 780. [Abstract] [Full Text] [PDF]

				M. M. Burg, M. C. Benedetto, R. Rosenberg, and R. Soufer Presurgical Depression Predicts Medical Morbidity 6 Months After Coronary Artery Bypass Graft Surgery Psychosom Med, January 1, 2003; 65(1): 111 - 118. [Abstract] [Full Text] [PDF]

				S. P. Sedlis, D. A. Morrison, J. D. Lorin, R. Esposito, G. Sethi, J. Sacks, W. Henderson, F. Grover, K. B. Ramanathan, D. Weiman, et al. Percutaneous coronary intervention versus coronary bypass graft surgery for diabetic patients with unstable angina and risk factors for adverse outcomes with bypass: outcome of diabetic patients in the AWESOME randomized trial and registry J. Am. Coll. Cardiol., November 6, 2002; 40(9): 1555 - 1566. [Abstract] [Full Text] [PDF]

				J. Steuer, P. Blomqvist, F. Granath, B. Rydh, A. Ekbom, U. de Faire, and E. Stahle Hospital readmission after coronary artery bypass grafting: are women doing worse? Ann. Thorac. Surg., May 1, 2002; 73(5): 1380 - 1386. [Abstract] [Full Text] [PDF]

				J. R. Handy Jr, A. I. Child, G. L. Grunkemeier, P. Fowler, J. W. Asaph, E. C. Douville, A. C. Tsen, and G. Y. Ott Hospital readmission after pulmonary resection: prevalence, patterns, and predisposing characteristics Ann. Thorac. Surg., December 1, 2001; 72(6): 1855 - 1860. [Abstract] [Full Text] [PDF]

				V. A. Ferraris, S. P. Ferraris, R. C. Harmon, and B. D. Evans Risk factors for early hospital readmission after cardiac operations J. Thorac. Cardiovasc. Surg., August 1, 2001; 122(2): 278 - 286. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Robert D. Stewart Sidney Levitsky Stephen J. Lahey Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stewart, R. D. Articles by Lahey, S. J. Search for Related Content PubMed PubMed Citation Articles by Stewart, R. D. Articles by Lahey, S. J.