Author's Accepted Manuscript
Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support Megan C. Maltby MSW, Maureen P. Flattery RN, MS, ANP, Brigid Burns MSW, Jeanne SalyerRN, PhD , Stephan Weinland PhD, Keyur B. Shah MD
http://www.jhltonline.org
PII: DOI: Reference:
S1053-2498(14)01089-4 http://dx.doi.org/10.1016/j.healun.2014.04.007 HEALUN5751
To appear in:
J Heart Lung Transplant
Cite this article as: Megan C. Maltby MSW, Maureen P. Flattery RN, MS, ANP, Brigid Burns MSW, Jeanne SalyerRN, PhD , Stephan Weinland PhD, Keyur B. Shah MD, Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support, J Heart Lung Transplant, http://dx.doi.org/ 10.1016/j.healun.2014.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support
Megan C. Maltby, MSW1, Maureen P. Flattery, RN, MS, ANP 2, Brigid Burns, MSW3, Jeanne Salyer, RN, PhD4, Stephan Weinland, PhD5 and Keyur B. Shah, MD 2
1
Department of Care Coordination, Virginia Commonwealth University Health System, Richmond, Virginia; 2Pauley Heart Center, Virginia Commonwealth University Health System, Richmond, Virginia; 3United Network for Organ Sharing, Richmond, Virginia; 4Virginia Commonwealth University School of Nursing, Richmond, Virginia; 5Department of Psychiatry, Virginia Commonwealth University Health System, Richmond, Virginia
Corresponding Author: Maureen Flattery, RN, MS, ANP PO Box 980204 Richmond, VA 23298 (P) 804-828-4571 (F) 804-828-7710 [email protected]
2 ABSTRACT Background: The Psychosocial Assessment of Candidates for Transplantation (PACT), developed to assess candidates for heart transplant (HT), has not been routinely used to assess left ventricular assist device (LVAD) candidacy. We examined the efficacy of the PACT to assess psychosocial outcomes in LVAD patients (pts.). Methods: We reviewed pts. implanted between June 2006 and April 2011 and retrospectively applied the PACT. We determined the accuracy of identifying social success with the PACT and revised domains to reflect criteria influencing social success for LVAD pts. Results: 48 pts. (72% men, 44% non-white, 50.4 yrs.) were divided into high and low scoring groups. Nine pts. with low PACT scores were falsely categorized as high risk while 4 with high scores had poor social outcomes. The score had a high positive predictive value (PPV=0.86) but low negative predictive value (NPV= 0.31). The PACT was revised (mPACT) to measure indicators that were identified to more closely affect LVAD outcome, such as social support and understanding of care requirements. The mPACT exhibited improved accuracy. A reclassification table was developed and the net reclassification index was 0.32. The percentage of patients incorrectly classified for social risk decreased from 27% with the PACT to 8% with the mPACT. Patients with higher mPACT scores had decreased 30-day readmission rates (26% vs. 67%, P=0.045) after device implantation. Conclusions: By emphasizing social support, psychological health, lifestyle factors and device understanding the mPACT showed improved performance in risk stratifying candidates for LVAD therapy. Prospective validation is warranted.
3 INTRODUCTION With the advent of continuous flow devices, use of implantable left ventricular assist devices (LVADs) has become an established therapy for patients with advanced heart failure.1 While much work has been done to determine psychosocial candidacy for heart transplantation.
2, 3, 4, 5
little research has
been conducted to assess psychosocial predictors of outcomes for patients requiring LVAD therapy.
2
Decisions whether to offer this therapy tend to be based on opinion and can be inconsistent. Objective assessment tools can be used to predict, plan for and influence treatment outcomes. Psychosocial factors impact patient outcome following LVAD implantation. There is evidence that supportive relationships influence participation in health behaviors in patients with chronic diseases. 6 ,7, 8 Current guidelines recommend investigating any history of psychiatric disorders, drug abuse, and other psychosocial issues that may impact a patient’s ability to understand and comply with LVAD care.1 Some recommend including cognitive testing as part of the standard evaluation in order to determine if candidates have the capacity to comply with recommendations. 9 In addition, a reliable social support system should be identified to assist the patient with care requirements. The Psychosocial Assessment of Candidates for Transplantation (PACT) was developed in 1989 to assess factors influencing psychosocial outcomes in transplant patients.2 It was originally developed to study the clinical judgment of the screeners. The PACT has been used to assist the decision making of transplant team members as to a patient’s candidacy.2 The intent was to incorporate its use into the evaluation process and to help the evaluating team to be consistent in decision making. Additional rating tools for psychosocial assessment of organ transplantation recipients have been developed but are not in widespread use.3, 4 To date, an objective assessment of the psychosocial risk factors of patients being evaluated for LVAD therapy has not been developed. The purpose of this study was to assess the utility of the PACT to measure psychosocial factors that may influence LVAD success. A further goal was to develop a new instrument that would more accurately evaluate these factors.
4 METHODS This was a single center, retrospective study that included all patients who underwent LVAD implantation between June 2006 and April 2011 and were eligible for discharge to home. Both bridge to transplant and destination therapy patients were included. Fifty of the 60 patients who received an LVAD between June, 2006 and April 2011 were eligible for discharge and therefore met criteria for inclusion in this study. Ten patients who received an LVAD during this time period either were transplanted during their index hospitalization or did not survive to discharge. One patient was excluded for incomplete data. One patient was incarcerated at the time of implant and was not discharged to independent living. Therefore, 48 patients were evaluated. Thirty-four of the 48 patients were considered transplant candidates. The Virginia Commonwealth University Institutional Review Board reviewed and approved this research. The PACT score The PACT consists of eight items measuring characteristics of psychosocial significance; support stability and availability, personality and psychological health, lifestyle factors, adherence, drug and alcohol use, and understanding of transplant and follow up. A clinical social worker, blinded to patient outcomes, retrospectively rated patient psychosocial functioning with the PACT using the evaluations obtained prior to implantation. The social worker did not know whether the patient received an LVAD as a bridge to transplant or for destination therapy. The PACT is graded using a 5 point Likert scale where a low score (0) is considered unacceptable and a high score (4) is considered excellent. The final score is not a total of the eight items; rather, it is the mode--the most commonly occurring score.2 This method was used because the authors wanted to be able to weight different factors that might play a role for individual patients. 2 In this study, patients were then divided into two groups based on their PACT scores. The low scoring group had scores less than 3 and the high scoring group had scores 3 or greater. Determining Social Success Social success was defined as having a reliable support system, discharged on schedule, and no social readmissions. Social readmission was defined as an admission for lack of placement or housing,
5 caregiver distress, or unsafe living conditions. Answers were assigned a value of 0 = “No” and 1 = “Yes”. Success was operationalized as an average score greater than 0.5, meaning that a majority of the team members felt the patient was successful. Raters, including two surgeons, two cardiologists, an LVAD coordinator, and social worker were asked, independently, “Was this patient a social success?” based on the above definition. Raters were given enough information to score the patient. This included etiology of heart failure and date of implant but no specific details about the patient so as not to bias the rater’s response. Raters were asked not to discuss their responses with other team members. Statistical Analysis and Creation of the mPACT The IBM SPSS Statistics, version 21 (Chicago, Illinois) software package was used for the analyses. Continuous variables were presented as means and standard deviation and categorical variables were presented as percentages. A two tailed Student’s t-test was used to compare continuous variables. Chi-square analysis was used to compare discrete variables. The ability of the PACT score to determine social success was analyzed. Using a PACT score cutoff 3 to identify a social success, we determined the sensitivity, specificity and positive and negative predictive values for the tool. For patients whose PACT score did not agree with the social outcome (PACT < 3 who were considered a social success [false negative] or PACT > 3 who were considered a social failure [false positive]), we performed a detailed chart review to determine the reason for misclassification. These findings informed revision of the PACT domains to better reflect psychosocial variables that influence social success with an LVAD. This new tool, the mPACT, was applied retrospectively to the same cohort and an optimal score cutoff was identified. In order to compare the difference in risk stratification of the two instruments, a net reclassification index was calculated.
6 RESULTS
Patient characteristics are presented in Table 1. Patients were predominantly Caucasian men who were supported by an LVAD for 278 days ± 196 days .Thirty-nine (81%) patients were considered a social success. The PACT low scoring group was younger and the high scoring group was more likely to have commercial insurance. There were no statistically significant differences in race, gender, etiology of heart failure or length of therapy between the two groups. The sensitivity of a PACT score > 3 to identify a social success was 0.77, and the specificity was 0.44. Positive predictive value of the score was 0.86 and negative predictive value was 0.31. Nine of the 13 patients with low PACT scores were falsely categorized as high risk for social failure. Reasons for low preoperative scores were history of non-adherence, poor understanding of their medical program, smoking, or concerns about caregiver ability. These issues, however, did not affect social outcome after device implantation. Four of the 35 patients with high PACT scores had poor social outcomes due to non-adherence or discharge delays secondary to anxiety and/or depression and lack of caregiver self-efficacy. A total of 27% of patients were incorrectly classified by the PACT. The mPACT
The PACT was revised (mPACT) to improve discrimination by measuring indicators that more closely affect LVAD outcome. To accomplish this, charts of all patients were reviewed, not just misclassified patients. Based on this review, emphasis was placed on social support, psychological health, lifestyle factors (e.g. ability to perform activities of daily living), and capability to understand care requirements. Areas of concern for heart transplant candidacy such as smoking were eliminated. Social support was then redefined to reflect the physical and psychological needs of a newly discharged LVAD patient The mPACT consists of four domains. Social support includes support stability and availability, psychological health consisting of psychopathology and personality factors, lifestyle factors including healthy lifestyle and independence with activities of daily living and capability to understand care
7 requirements consists of relevant knowledge, recognition of disease process, understanding of treatment options including LVAD therapy, and adherence. (Figure1). We considered factors that are known to affect LVAD outcomes. Within the literature, these factors are limited to medical complications, many of which require hospitalization.10, 11, 12 A psychosocial evaluation for patients with chronic health problems often incorporates factors such as support, caregiver ability, and barriers to adherence.13 Therefore in the re-conceptualization of factors influencing success or failure in our patient population we specifically addressed these factors in the mPACT. Questions were re-worded to make them easier for the clinician to rate. Additionally, overlapping questions were compressed into a single question. For instance, the stability and availability of caregiver support is rated in two separate questions using the PACT. This was compressed into one question in the mPACT. The scores were evenly weighted. Each individual indicator was rated 0, 1, or 2. (none, potential problem, or acceptable) Unlike the PACT, the mPACT was scored by summing the responses in each of the four domains. An mPACT score could range from 0 to 20. We felt that a sum of scores would be easier than determining the frequency of a particular response. Higher scores reflect a greater likelihood of social success. Total mPACT scores in this study ranged from 12 to 20. Evaluation of sensitivity, specificity, negative and positive predictive values determined that the optimal score, indicating social success, was 17 (Table 2). Patients above and below the cutoff score of 17 were similar in regards to age, gender, etiology of heart failure, ethnicity and duration of therapy. More patients in the high mPACT score group had commercial insurance, but this was only a trend. (Table 1) Net Reclassification Index (NRI) can evaluate the benefit of adding a new marker to an established set of risk factors and assess the improvement in prediction of a new strategy compared to an old one. The NRI is calculated by adding together the differences between “events” reclassified either higher or lower and “nonevents” reclassified higher or lower and dividing by the number of events or
8 nonevents.14 In our study, events were “social successes” and nonevents were “failures”. Eight of nine patients who were falsely categorized as “failures” by PACT were reclassified as “successes” by mPACT. Two of the five patients falsely categorized as “successes” by PACT were reclassified as “failures” by mPACT giving us a net reclassification index of 0.32. The percentage of patients incorrectly classified for social risk decreased from 27% with the PACT to 8% with the mPACT. We then examined the components of the mPACT. There were no differences between groups in the scores for social support, psychological health and lifestyle factors. Patients with scores 17 on the mPACT scored significantly higher on the “Capability to Understand the Device” section. (Table 3) More patients in the lower scoring group were readmitted within thirty days after discharge from the hospital. (Table 1) Although most readmissions in both groups were for medical reasons (syncope, arrhythmia, neurological events, fever, and right heart failure) two of the readmissions in the lower scoring group were for social reasons. Both readmissions were from the inpatient rehabilitation unit and were due to family concerns regarding their own ability to provide care. DISCUSSION The findings of this study suggest that compared to the PACT score, the newly created mPACT score more accurately identified patients who achieved “social success” after LVAD implantation. Moreover, patients with higher mPACT scores had lower hospital readmission rates. This revised instrument may have a role in the preoperative risk stratification for patients considering LVAD therapy and warrants prospective validation. The utility of standardized instruments for the evaluation of candidacy lies in the ability to afford standardized data which, when used by LVAD programs may provide the opportunity to make comparisons across patients and centers. Although the “gestalt” of an experienced practitioner may result in similar recommendations, an ability to quantify will provide health care professionals with objective data.
9 Technical improvements in LVAD therapy, prolonged waiting time for donor organs and changes to Medicare guidelines have all contributed to the increasing use of mechanical support, both as a bridge to transplantation and destination therapy. As more patients are referred to LVAD centers, those centers often struggle with predicting the impact of non-medical factors, such as degree of family support, ability to provide self-care, and impact of previous non-adherence on LVAD outcome. To date, a way to objectively and consistently evaluate psychosocial issues that may affect patient outcomes following LVAD implantation has not been developed. Consequently, evaluations are often based on opinion. Current instruments assess psychosocial factors in a standardized fashion in order to identify transplant candidates who are at risk for negative outcomes. 2, 3, 4 However, factors that may adversely affect transplant candidacy may not necessarily apply to the LVAD population. Unlike findings in heart transplantation recipients,15, 16, 17 we did not find any research addressing an association between smoking and LVAD outcomes. Similarly, although it has been recommended to apply the same standard for medical adherence to LVAD patients as is used for heart transplant candidates, there is little in the literature concerning the relationship between a history of non-adherence and specific LVAD outcomes. 18, 19, 20
Smoking has long been considered a contraindication to heart transplantation due to its association
with decreased survival and increased incidence of both coronary vasculopathy and malignancies.15, 16, 17 Smoking, either before or after transplantation has not been studied as an influence on psychosocial outcomes, rather its deleterious health effects are the reason it is considered a contraindication. Patients in our study who smoked received lower scores on the PACT than they did on the mPACT. They were not considered transplant candidates until they had demonstrated a six month abstinence from smoking as assessed by urine cotinine. Patients were offered both pharmacologic and behavioral therapy in support of cessation efforts. This however was not the primary concern for revising the instrument. The purpose of this initiative (revision of the PACT) was to examine factors that are predictive of a successful transition to home with an LVAD.
10 Farmer, Grady et al in a study of demographic, psychosocial and behavioral factors associated with survival after transplantation found that poor adherence to the pre-transplantation medical regimen was related to poor adherence to the post-transplantation regimen and consequently, to poor outcomes.21 Transplant patients, although required to take medication do not have the same reminder of their clinical condition as LVAD patients. Patients with an LVAD are constantly reminded of their medical vulnerability by their driveline and device components. Perhaps these are stronger cues for the need to adhere to a prescribed medical regimen than prescriptions and appointments. Patients in the higher scoring group tended to have commercial insurance. These results are similar to findings from a review of the UNOS database that revealed that post transplantation survival differed by type of insurance.22 Medicare and Medicaid patients had significantly lower ten year survival following transplantation than patients with commercial insurance.22 This may reflect the patient’s socioeconomic status which may influence social success after LVAD implantation. Further study to clarify the influence of socioeconomic resources on the attributes essential to a successful post-implantation trajectory is warranted. Patients with an mPACT score 17 had higher scores in the “Capability to Understand Care Requirements” section. This section did not examine actual device knowledge since most patients had not yet received LVAD training; rather it examined a patient’s understanding of current medical condition and need for advanced therapy. Patients with an mPACT score < 17 had higher 30-day re-hospitalization rates. The Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure study suggested that readmissions for LVAD patients are frequent.10 Studies to investigate readmission following LVAD implantation have, however, focused on medically indicated readmission. Forest et al reported that 56 of 92 patients were readmitted during device therapy, including 25 who were readmitted within 30 days of initial discharge.11 The most common etiology for readmission was gastrointestinal bleeding followed by infection, device alarms and pain. A retrospective analysis of 115 patients implanted with a continuous flow LVAD by Hasin and colleagues revealed an overall readmission rate of 1.64 per patient year of
11 follow up and were caused, as in the previous study, by bleeding, cardiac indications, infection and thrombosis.12 The influence of poor social support on readmission has not been examined. Identifying risk factors prospectively that may adversely influence social success may make it possible to develop strategies to strengthen weak social support, prevent non-medical readmissions, decrease cost and improve quality of life. This evaluation is only part of a comprehensive evaluation of candidacy for bridge to transplant or destination therapy and does not replace evaluation of lifestyle and economic factors that may impact adjustment/success. Specifically, the mPACT evaluation should not be used to exclude a candidate, but rather to identify potential psychosocial problems that will need to be addressed. Future studies might target only patients who receive destination therapy. LIMITATIONS This was a retrospective analysis and included only patients who received a device. Patients who were turned down for device therapy during this time period were not included in this analysis so there was selection bias. The cohort studied was a small number of patients with few events. Findings such as the thirty day readmission rate could be the result of a Type I () error. The mPACT must therefore be validated in another, larger LVAD cohort to define its utility. Perceptions of events by team members may have affected social success scoring. Additionally, this was a single center analysis. Practices and attitudes of the team can influence both outcomes and patient selection. CONCLUSIONS While the PACT score predicted social success after implantation of an LVAD, it falsely categorized many patients as high risk. By emphasizing social support, psychological health, lifestyle factors and device understanding the mPACT showed improved performance in risk stratifying candidates for LVAD therapy. Prospective validation is warranted.
12 DISCLOSURES Megan Maltby, MSW, Maureen Flattery, RN, MS, ANP, Brigid Burns, MSW, Jeanne Salyer, RN, PhD, and Stephan Weinland, PhD have no disclosures. Keyur Shah, MD has received institutional grants from Thoratec, Corp. ACKNOWLEDGEMENTS The authors wish to thank Anit Mankad, MD, Alice Barclay, MSW, and Mary Ellen Olbrisch PhD for their assistance in completing this project.
13 REFERENCES 1. Slaughter MS, Pagani FD, Rogers JG et al. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant 2010; 29: S2-S39. 2. Olbrisch ME, Levenson JL & Hamer R. The PACT: A rating scale for the study of clinical decision making in psychosocial screening of organ transplant candidates. Clin Transplant 1989; 3:164-9. 3. Maldonado JR, Dubois H, David E et al. The Stanford integrated psychosocial assessment for transplantation (SIPAT): A new tool for the psychosocial evaluation of pre-transplant candidates. Psychosomatics 2012: 123-32. 4. Twillman RK, Manetto C, Wellisch DK & Wolcott DL. The Transplant Evaluation Rating Scale. Psychosomatics 1993; 34: 144-53. 5. Olbrisch ME, Benedict SM, Ashe K & Levenson JL. Psychological assessment and care of organ transplant patients. J Consult Clin Psychol. 2002; 70: 771-83.6.. Salyer J, Schubert CM & Chiaranai C. Supportive relationahips, self-care confidence, and heart failure self-care. J Cardiovasc Nurs 2012; 27: 384-93. 7. Sebern M & Riegel B. Contributions of supportive relationships to heart failure self-care. Eur J Cardiovasc Nurs 2009; 8: 97-104. 8.Calvillo-King L, Arnold D, Eubank KJ et al. Impact of factors on risk of readmission or mortality in pneumonia and heart failure: systematic review. J Gen Intern Med 2013; 28: 269-82. 9. Miller LW & Guglin M. Patient selection for ventricular assist devices: A moving target. J Amer Coll Cardiol 2013; 61: 1209-21. 10 . Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of left ventricular assist device for end stage heart failure. N Eng J Med 2001; 3345: 1435-43. 11 . Forest SJ, Bello R, Friedmann P et al. Readmissions after ventricular assist device: etiologies, patterns and days out of hospital. Ann Thorac Surg 2012; 95: 1276-81.
14 12. Hasin T, Marmor Y, Kremers W et al. Readmissions after implantation of axial flow ventricular assist device. J Am Coll Cardiol 2013; 61: 153-63 13. Kocaman N, Kutlu Y, Ozkan M & Ozkan S. Predictors of psychosocial adjustment in people with physical disease. J Clin Nurs 2007;16: 6-16. 14. Pepe MS. Problems with risk reclassification methods for evaluating prediction models. Am J Epidemiol 2011; 173: 1327-35. 15. Arora S, Aukrust P, Andreassen A et al. The prognostic importance of modifiable risk factors after heart transplantation. Am Heart J 2009; 158-431. 16. Sanchez-Lazaro IJ, Almenar L, Martinez-Dolz L et al. Effect of hypertension, diabetes, and smoking on development of renal dysfunction after heart transplantation Transplant Proc 2008; 40: 3049-50. 17. Yagdi T, Sharples L, Tsui S et al. Malignancy after heart transplantation: analysis of 24 year experience at a single center. J Card Surg 2009; 24: 572-9. 18. Eschelman AK, Mason S, Nemeh H & Williams C. LVAD destination therapy: applying what we know about psychiatric evaluation and management from cardiac failure and transplant. Heart Fail Rev 2009; 14: 21-8. 19. Vitale CA, Chandekar R, Rodgers PE, Pagani FD & Malani PN. A call for guidance in the use of left ventricular assist devices in older adults. J Am Geriatr Soc 2012; 60: 145-50. 20. Dew MA, Kormos RL, Roth LH, Murali S, DiMartini A & Griffith BP. Early post-transplant medical compliance and mental health predicts physical morbidity and mortality 1 to 3 years after heart transplantation. J Heart Lung Transplant 1999; 18: 549-562. 21. Farmer SA, Grady KL, Wang E, McGee EC, Cotts WG & McCarthy PM. Demographic, psychosocial, and behavioral factors Associated with survival after heart transplantation. Ann Thorac Surg 2013; 95: 876-83. 22. Allen JG, Weiss ES, Arnaoutakis GJ, et al. Insurance and education predict long-term survival after orthotopic heart transplantation in the United States. J Heart Lung Transplant 2012; 31: 52-60.
15 Figure Legend. Figure 1. The revised scoring instrument, re-written to incorporate indicators that more closely reflect LVAD outcome. These include social support which explores support stability and availability, psychological health which consists of psychopathology and personality factors, lifestyle factors including healthy lifestyle and independence with activities of daily living and capability to understand care requirements consists of relevant knowledge, understanding of disease process and adherence.
16 Table 1. Patient Characteristics Total Cohort (n=48) Age at implant 50.6±13.7 (years, mean ± SD) Commercial 15 (31%) insurance Non-white race 27 (56%) Female gender 25% Non ischemic 60% etiology Length of stay 41.6±21.3 (days, mean ± SD) Duration of 278±196 therapy (days, mean ± SD) Destination 14 (29%) Therapy Social success 39 (83%) 30 day 15 (31%) readmission SD=standard deviation
High mPACT Score (17) n=42 50.8±12.8
p value
0.03
Low mPACT Score (
Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support Megan C. Maltby MSW, Maureen P. Flattery RN, MS, ANP, Brigid Burns MSW, Jeanne SalyerRN, PhD , Stephan Weinland PhD, Keyur B. Shah MD
http://www.jhltonline.org
PII: DOI: Reference:
S1053-2498(14)01089-4 http://dx.doi.org/10.1016/j.healun.2014.04.007 HEALUN5751
To appear in:
J Heart Lung Transplant
Cite this article as: Megan C. Maltby MSW, Maureen P. Flattery RN, MS, ANP, Brigid Burns MSW, Jeanne SalyerRN, PhD , Stephan Weinland PhD, Keyur B. Shah MD, Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support, J Heart Lung Transplant, http://dx.doi.org/ 10.1016/j.healun.2014.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Psychosocial Assessment of Candidates and Risk Classification of Patients Considered for Durable Mechanical Circulatory Support
Megan C. Maltby, MSW1, Maureen P. Flattery, RN, MS, ANP 2, Brigid Burns, MSW3, Jeanne Salyer, RN, PhD4, Stephan Weinland, PhD5 and Keyur B. Shah, MD 2
1
Department of Care Coordination, Virginia Commonwealth University Health System, Richmond, Virginia; 2Pauley Heart Center, Virginia Commonwealth University Health System, Richmond, Virginia; 3United Network for Organ Sharing, Richmond, Virginia; 4Virginia Commonwealth University School of Nursing, Richmond, Virginia; 5Department of Psychiatry, Virginia Commonwealth University Health System, Richmond, Virginia
Corresponding Author: Maureen Flattery, RN, MS, ANP PO Box 980204 Richmond, VA 23298 (P) 804-828-4571 (F) 804-828-7710 [email protected]
2 ABSTRACT Background: The Psychosocial Assessment of Candidates for Transplantation (PACT), developed to assess candidates for heart transplant (HT), has not been routinely used to assess left ventricular assist device (LVAD) candidacy. We examined the efficacy of the PACT to assess psychosocial outcomes in LVAD patients (pts.). Methods: We reviewed pts. implanted between June 2006 and April 2011 and retrospectively applied the PACT. We determined the accuracy of identifying social success with the PACT and revised domains to reflect criteria influencing social success for LVAD pts. Results: 48 pts. (72% men, 44% non-white, 50.4 yrs.) were divided into high and low scoring groups. Nine pts. with low PACT scores were falsely categorized as high risk while 4 with high scores had poor social outcomes. The score had a high positive predictive value (PPV=0.86) but low negative predictive value (NPV= 0.31). The PACT was revised (mPACT) to measure indicators that were identified to more closely affect LVAD outcome, such as social support and understanding of care requirements. The mPACT exhibited improved accuracy. A reclassification table was developed and the net reclassification index was 0.32. The percentage of patients incorrectly classified for social risk decreased from 27% with the PACT to 8% with the mPACT. Patients with higher mPACT scores had decreased 30-day readmission rates (26% vs. 67%, P=0.045) after device implantation. Conclusions: By emphasizing social support, psychological health, lifestyle factors and device understanding the mPACT showed improved performance in risk stratifying candidates for LVAD therapy. Prospective validation is warranted.
3 INTRODUCTION With the advent of continuous flow devices, use of implantable left ventricular assist devices (LVADs) has become an established therapy for patients with advanced heart failure.1 While much work has been done to determine psychosocial candidacy for heart transplantation.
2, 3, 4, 5
little research has
been conducted to assess psychosocial predictors of outcomes for patients requiring LVAD therapy.
2
Decisions whether to offer this therapy tend to be based on opinion and can be inconsistent. Objective assessment tools can be used to predict, plan for and influence treatment outcomes. Psychosocial factors impact patient outcome following LVAD implantation. There is evidence that supportive relationships influence participation in health behaviors in patients with chronic diseases. 6 ,7, 8 Current guidelines recommend investigating any history of psychiatric disorders, drug abuse, and other psychosocial issues that may impact a patient’s ability to understand and comply with LVAD care.1 Some recommend including cognitive testing as part of the standard evaluation in order to determine if candidates have the capacity to comply with recommendations. 9 In addition, a reliable social support system should be identified to assist the patient with care requirements. The Psychosocial Assessment of Candidates for Transplantation (PACT) was developed in 1989 to assess factors influencing psychosocial outcomes in transplant patients.2 It was originally developed to study the clinical judgment of the screeners. The PACT has been used to assist the decision making of transplant team members as to a patient’s candidacy.2 The intent was to incorporate its use into the evaluation process and to help the evaluating team to be consistent in decision making. Additional rating tools for psychosocial assessment of organ transplantation recipients have been developed but are not in widespread use.3, 4 To date, an objective assessment of the psychosocial risk factors of patients being evaluated for LVAD therapy has not been developed. The purpose of this study was to assess the utility of the PACT to measure psychosocial factors that may influence LVAD success. A further goal was to develop a new instrument that would more accurately evaluate these factors.
4 METHODS This was a single center, retrospective study that included all patients who underwent LVAD implantation between June 2006 and April 2011 and were eligible for discharge to home. Both bridge to transplant and destination therapy patients were included. Fifty of the 60 patients who received an LVAD between June, 2006 and April 2011 were eligible for discharge and therefore met criteria for inclusion in this study. Ten patients who received an LVAD during this time period either were transplanted during their index hospitalization or did not survive to discharge. One patient was excluded for incomplete data. One patient was incarcerated at the time of implant and was not discharged to independent living. Therefore, 48 patients were evaluated. Thirty-four of the 48 patients were considered transplant candidates. The Virginia Commonwealth University Institutional Review Board reviewed and approved this research. The PACT score The PACT consists of eight items measuring characteristics of psychosocial significance; support stability and availability, personality and psychological health, lifestyle factors, adherence, drug and alcohol use, and understanding of transplant and follow up. A clinical social worker, blinded to patient outcomes, retrospectively rated patient psychosocial functioning with the PACT using the evaluations obtained prior to implantation. The social worker did not know whether the patient received an LVAD as a bridge to transplant or for destination therapy. The PACT is graded using a 5 point Likert scale where a low score (0) is considered unacceptable and a high score (4) is considered excellent. The final score is not a total of the eight items; rather, it is the mode--the most commonly occurring score.2 This method was used because the authors wanted to be able to weight different factors that might play a role for individual patients. 2 In this study, patients were then divided into two groups based on their PACT scores. The low scoring group had scores less than 3 and the high scoring group had scores 3 or greater. Determining Social Success Social success was defined as having a reliable support system, discharged on schedule, and no social readmissions. Social readmission was defined as an admission for lack of placement or housing,
5 caregiver distress, or unsafe living conditions. Answers were assigned a value of 0 = “No” and 1 = “Yes”. Success was operationalized as an average score greater than 0.5, meaning that a majority of the team members felt the patient was successful. Raters, including two surgeons, two cardiologists, an LVAD coordinator, and social worker were asked, independently, “Was this patient a social success?” based on the above definition. Raters were given enough information to score the patient. This included etiology of heart failure and date of implant but no specific details about the patient so as not to bias the rater’s response. Raters were asked not to discuss their responses with other team members. Statistical Analysis and Creation of the mPACT The IBM SPSS Statistics, version 21 (Chicago, Illinois) software package was used for the analyses. Continuous variables were presented as means and standard deviation and categorical variables were presented as percentages. A two tailed Student’s t-test was used to compare continuous variables. Chi-square analysis was used to compare discrete variables. The ability of the PACT score to determine social success was analyzed. Using a PACT score cutoff 3 to identify a social success, we determined the sensitivity, specificity and positive and negative predictive values for the tool. For patients whose PACT score did not agree with the social outcome (PACT < 3 who were considered a social success [false negative] or PACT > 3 who were considered a social failure [false positive]), we performed a detailed chart review to determine the reason for misclassification. These findings informed revision of the PACT domains to better reflect psychosocial variables that influence social success with an LVAD. This new tool, the mPACT, was applied retrospectively to the same cohort and an optimal score cutoff was identified. In order to compare the difference in risk stratification of the two instruments, a net reclassification index was calculated.
6 RESULTS
Patient characteristics are presented in Table 1. Patients were predominantly Caucasian men who were supported by an LVAD for 278 days ± 196 days .Thirty-nine (81%) patients were considered a social success. The PACT low scoring group was younger and the high scoring group was more likely to have commercial insurance. There were no statistically significant differences in race, gender, etiology of heart failure or length of therapy between the two groups. The sensitivity of a PACT score > 3 to identify a social success was 0.77, and the specificity was 0.44. Positive predictive value of the score was 0.86 and negative predictive value was 0.31. Nine of the 13 patients with low PACT scores were falsely categorized as high risk for social failure. Reasons for low preoperative scores were history of non-adherence, poor understanding of their medical program, smoking, or concerns about caregiver ability. These issues, however, did not affect social outcome after device implantation. Four of the 35 patients with high PACT scores had poor social outcomes due to non-adherence or discharge delays secondary to anxiety and/or depression and lack of caregiver self-efficacy. A total of 27% of patients were incorrectly classified by the PACT. The mPACT
The PACT was revised (mPACT) to improve discrimination by measuring indicators that more closely affect LVAD outcome. To accomplish this, charts of all patients were reviewed, not just misclassified patients. Based on this review, emphasis was placed on social support, psychological health, lifestyle factors (e.g. ability to perform activities of daily living), and capability to understand care requirements. Areas of concern for heart transplant candidacy such as smoking were eliminated. Social support was then redefined to reflect the physical and psychological needs of a newly discharged LVAD patient The mPACT consists of four domains. Social support includes support stability and availability, psychological health consisting of psychopathology and personality factors, lifestyle factors including healthy lifestyle and independence with activities of daily living and capability to understand care
7 requirements consists of relevant knowledge, recognition of disease process, understanding of treatment options including LVAD therapy, and adherence. (Figure1). We considered factors that are known to affect LVAD outcomes. Within the literature, these factors are limited to medical complications, many of which require hospitalization.10, 11, 12 A psychosocial evaluation for patients with chronic health problems often incorporates factors such as support, caregiver ability, and barriers to adherence.13 Therefore in the re-conceptualization of factors influencing success or failure in our patient population we specifically addressed these factors in the mPACT. Questions were re-worded to make them easier for the clinician to rate. Additionally, overlapping questions were compressed into a single question. For instance, the stability and availability of caregiver support is rated in two separate questions using the PACT. This was compressed into one question in the mPACT. The scores were evenly weighted. Each individual indicator was rated 0, 1, or 2. (none, potential problem, or acceptable) Unlike the PACT, the mPACT was scored by summing the responses in each of the four domains. An mPACT score could range from 0 to 20. We felt that a sum of scores would be easier than determining the frequency of a particular response. Higher scores reflect a greater likelihood of social success. Total mPACT scores in this study ranged from 12 to 20. Evaluation of sensitivity, specificity, negative and positive predictive values determined that the optimal score, indicating social success, was 17 (Table 2). Patients above and below the cutoff score of 17 were similar in regards to age, gender, etiology of heart failure, ethnicity and duration of therapy. More patients in the high mPACT score group had commercial insurance, but this was only a trend. (Table 1) Net Reclassification Index (NRI) can evaluate the benefit of adding a new marker to an established set of risk factors and assess the improvement in prediction of a new strategy compared to an old one. The NRI is calculated by adding together the differences between “events” reclassified either higher or lower and “nonevents” reclassified higher or lower and dividing by the number of events or
8 nonevents.14 In our study, events were “social successes” and nonevents were “failures”. Eight of nine patients who were falsely categorized as “failures” by PACT were reclassified as “successes” by mPACT. Two of the five patients falsely categorized as “successes” by PACT were reclassified as “failures” by mPACT giving us a net reclassification index of 0.32. The percentage of patients incorrectly classified for social risk decreased from 27% with the PACT to 8% with the mPACT. We then examined the components of the mPACT. There were no differences between groups in the scores for social support, psychological health and lifestyle factors. Patients with scores 17 on the mPACT scored significantly higher on the “Capability to Understand the Device” section. (Table 3) More patients in the lower scoring group were readmitted within thirty days after discharge from the hospital. (Table 1) Although most readmissions in both groups were for medical reasons (syncope, arrhythmia, neurological events, fever, and right heart failure) two of the readmissions in the lower scoring group were for social reasons. Both readmissions were from the inpatient rehabilitation unit and were due to family concerns regarding their own ability to provide care. DISCUSSION The findings of this study suggest that compared to the PACT score, the newly created mPACT score more accurately identified patients who achieved “social success” after LVAD implantation. Moreover, patients with higher mPACT scores had lower hospital readmission rates. This revised instrument may have a role in the preoperative risk stratification for patients considering LVAD therapy and warrants prospective validation. The utility of standardized instruments for the evaluation of candidacy lies in the ability to afford standardized data which, when used by LVAD programs may provide the opportunity to make comparisons across patients and centers. Although the “gestalt” of an experienced practitioner may result in similar recommendations, an ability to quantify will provide health care professionals with objective data.
9 Technical improvements in LVAD therapy, prolonged waiting time for donor organs and changes to Medicare guidelines have all contributed to the increasing use of mechanical support, both as a bridge to transplantation and destination therapy. As more patients are referred to LVAD centers, those centers often struggle with predicting the impact of non-medical factors, such as degree of family support, ability to provide self-care, and impact of previous non-adherence on LVAD outcome. To date, a way to objectively and consistently evaluate psychosocial issues that may affect patient outcomes following LVAD implantation has not been developed. Consequently, evaluations are often based on opinion. Current instruments assess psychosocial factors in a standardized fashion in order to identify transplant candidates who are at risk for negative outcomes. 2, 3, 4 However, factors that may adversely affect transplant candidacy may not necessarily apply to the LVAD population. Unlike findings in heart transplantation recipients,15, 16, 17 we did not find any research addressing an association between smoking and LVAD outcomes. Similarly, although it has been recommended to apply the same standard for medical adherence to LVAD patients as is used for heart transplant candidates, there is little in the literature concerning the relationship between a history of non-adherence and specific LVAD outcomes. 18, 19, 20
Smoking has long been considered a contraindication to heart transplantation due to its association
with decreased survival and increased incidence of both coronary vasculopathy and malignancies.15, 16, 17 Smoking, either before or after transplantation has not been studied as an influence on psychosocial outcomes, rather its deleterious health effects are the reason it is considered a contraindication. Patients in our study who smoked received lower scores on the PACT than they did on the mPACT. They were not considered transplant candidates until they had demonstrated a six month abstinence from smoking as assessed by urine cotinine. Patients were offered both pharmacologic and behavioral therapy in support of cessation efforts. This however was not the primary concern for revising the instrument. The purpose of this initiative (revision of the PACT) was to examine factors that are predictive of a successful transition to home with an LVAD.
10 Farmer, Grady et al in a study of demographic, psychosocial and behavioral factors associated with survival after transplantation found that poor adherence to the pre-transplantation medical regimen was related to poor adherence to the post-transplantation regimen and consequently, to poor outcomes.21 Transplant patients, although required to take medication do not have the same reminder of their clinical condition as LVAD patients. Patients with an LVAD are constantly reminded of their medical vulnerability by their driveline and device components. Perhaps these are stronger cues for the need to adhere to a prescribed medical regimen than prescriptions and appointments. Patients in the higher scoring group tended to have commercial insurance. These results are similar to findings from a review of the UNOS database that revealed that post transplantation survival differed by type of insurance.22 Medicare and Medicaid patients had significantly lower ten year survival following transplantation than patients with commercial insurance.22 This may reflect the patient’s socioeconomic status which may influence social success after LVAD implantation. Further study to clarify the influence of socioeconomic resources on the attributes essential to a successful post-implantation trajectory is warranted. Patients with an mPACT score 17 had higher scores in the “Capability to Understand Care Requirements” section. This section did not examine actual device knowledge since most patients had not yet received LVAD training; rather it examined a patient’s understanding of current medical condition and need for advanced therapy. Patients with an mPACT score < 17 had higher 30-day re-hospitalization rates. The Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure study suggested that readmissions for LVAD patients are frequent.10 Studies to investigate readmission following LVAD implantation have, however, focused on medically indicated readmission. Forest et al reported that 56 of 92 patients were readmitted during device therapy, including 25 who were readmitted within 30 days of initial discharge.11 The most common etiology for readmission was gastrointestinal bleeding followed by infection, device alarms and pain. A retrospective analysis of 115 patients implanted with a continuous flow LVAD by Hasin and colleagues revealed an overall readmission rate of 1.64 per patient year of
11 follow up and were caused, as in the previous study, by bleeding, cardiac indications, infection and thrombosis.12 The influence of poor social support on readmission has not been examined. Identifying risk factors prospectively that may adversely influence social success may make it possible to develop strategies to strengthen weak social support, prevent non-medical readmissions, decrease cost and improve quality of life. This evaluation is only part of a comprehensive evaluation of candidacy for bridge to transplant or destination therapy and does not replace evaluation of lifestyle and economic factors that may impact adjustment/success. Specifically, the mPACT evaluation should not be used to exclude a candidate, but rather to identify potential psychosocial problems that will need to be addressed. Future studies might target only patients who receive destination therapy. LIMITATIONS This was a retrospective analysis and included only patients who received a device. Patients who were turned down for device therapy during this time period were not included in this analysis so there was selection bias. The cohort studied was a small number of patients with few events. Findings such as the thirty day readmission rate could be the result of a Type I () error. The mPACT must therefore be validated in another, larger LVAD cohort to define its utility. Perceptions of events by team members may have affected social success scoring. Additionally, this was a single center analysis. Practices and attitudes of the team can influence both outcomes and patient selection. CONCLUSIONS While the PACT score predicted social success after implantation of an LVAD, it falsely categorized many patients as high risk. By emphasizing social support, psychological health, lifestyle factors and device understanding the mPACT showed improved performance in risk stratifying candidates for LVAD therapy. Prospective validation is warranted.
12 DISCLOSURES Megan Maltby, MSW, Maureen Flattery, RN, MS, ANP, Brigid Burns, MSW, Jeanne Salyer, RN, PhD, and Stephan Weinland, PhD have no disclosures. Keyur Shah, MD has received institutional grants from Thoratec, Corp. ACKNOWLEDGEMENTS The authors wish to thank Anit Mankad, MD, Alice Barclay, MSW, and Mary Ellen Olbrisch PhD for their assistance in completing this project.
13 REFERENCES 1. Slaughter MS, Pagani FD, Rogers JG et al. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant 2010; 29: S2-S39. 2. Olbrisch ME, Levenson JL & Hamer R. The PACT: A rating scale for the study of clinical decision making in psychosocial screening of organ transplant candidates. Clin Transplant 1989; 3:164-9. 3. Maldonado JR, Dubois H, David E et al. The Stanford integrated psychosocial assessment for transplantation (SIPAT): A new tool for the psychosocial evaluation of pre-transplant candidates. Psychosomatics 2012: 123-32. 4. Twillman RK, Manetto C, Wellisch DK & Wolcott DL. The Transplant Evaluation Rating Scale. Psychosomatics 1993; 34: 144-53. 5. Olbrisch ME, Benedict SM, Ashe K & Levenson JL. Psychological assessment and care of organ transplant patients. J Consult Clin Psychol. 2002; 70: 771-83.6.. Salyer J, Schubert CM & Chiaranai C. Supportive relationahips, self-care confidence, and heart failure self-care. J Cardiovasc Nurs 2012; 27: 384-93. 7. Sebern M & Riegel B. Contributions of supportive relationships to heart failure self-care. Eur J Cardiovasc Nurs 2009; 8: 97-104. 8.Calvillo-King L, Arnold D, Eubank KJ et al. Impact of factors on risk of readmission or mortality in pneumonia and heart failure: systematic review. J Gen Intern Med 2013; 28: 269-82. 9. Miller LW & Guglin M. Patient selection for ventricular assist devices: A moving target. J Amer Coll Cardiol 2013; 61: 1209-21. 10 . Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of left ventricular assist device for end stage heart failure. N Eng J Med 2001; 3345: 1435-43. 11 . Forest SJ, Bello R, Friedmann P et al. Readmissions after ventricular assist device: etiologies, patterns and days out of hospital. Ann Thorac Surg 2012; 95: 1276-81.
14 12. Hasin T, Marmor Y, Kremers W et al. Readmissions after implantation of axial flow ventricular assist device. J Am Coll Cardiol 2013; 61: 153-63 13. Kocaman N, Kutlu Y, Ozkan M & Ozkan S. Predictors of psychosocial adjustment in people with physical disease. J Clin Nurs 2007;16: 6-16. 14. Pepe MS. Problems with risk reclassification methods for evaluating prediction models. Am J Epidemiol 2011; 173: 1327-35. 15. Arora S, Aukrust P, Andreassen A et al. The prognostic importance of modifiable risk factors after heart transplantation. Am Heart J 2009; 158-431. 16. Sanchez-Lazaro IJ, Almenar L, Martinez-Dolz L et al. Effect of hypertension, diabetes, and smoking on development of renal dysfunction after heart transplantation Transplant Proc 2008; 40: 3049-50. 17. Yagdi T, Sharples L, Tsui S et al. Malignancy after heart transplantation: analysis of 24 year experience at a single center. J Card Surg 2009; 24: 572-9. 18. Eschelman AK, Mason S, Nemeh H & Williams C. LVAD destination therapy: applying what we know about psychiatric evaluation and management from cardiac failure and transplant. Heart Fail Rev 2009; 14: 21-8. 19. Vitale CA, Chandekar R, Rodgers PE, Pagani FD & Malani PN. A call for guidance in the use of left ventricular assist devices in older adults. J Am Geriatr Soc 2012; 60: 145-50. 20. Dew MA, Kormos RL, Roth LH, Murali S, DiMartini A & Griffith BP. Early post-transplant medical compliance and mental health predicts physical morbidity and mortality 1 to 3 years after heart transplantation. J Heart Lung Transplant 1999; 18: 549-562. 21. Farmer SA, Grady KL, Wang E, McGee EC, Cotts WG & McCarthy PM. Demographic, psychosocial, and behavioral factors Associated with survival after heart transplantation. Ann Thorac Surg 2013; 95: 876-83. 22. Allen JG, Weiss ES, Arnaoutakis GJ, et al. Insurance and education predict long-term survival after orthotopic heart transplantation in the United States. J Heart Lung Transplant 2012; 31: 52-60.
15 Figure Legend. Figure 1. The revised scoring instrument, re-written to incorporate indicators that more closely reflect LVAD outcome. These include social support which explores support stability and availability, psychological health which consists of psychopathology and personality factors, lifestyle factors including healthy lifestyle and independence with activities of daily living and capability to understand care requirements consists of relevant knowledge, understanding of disease process and adherence.
16 Table 1. Patient Characteristics Total Cohort (n=48) Age at implant 50.6±13.7 (years, mean ± SD) Commercial 15 (31%) insurance Non-white race 27 (56%) Female gender 25% Non ischemic 60% etiology Length of stay 41.6±21.3 (days, mean ± SD) Duration of 278±196 therapy (days, mean ± SD) Destination 14 (29%) Therapy Social success 39 (83%) 30 day 15 (31%) readmission SD=standard deviation
High mPACT Score (17) n=42 50.8±12.8
p value
0.03
Low mPACT Score (
