Surgical Treatment Outcomes for Patients with Psychiatric Disorders (STOPP)
Laurel Anne Copeland PhD MPH BS
Central Texas Veterans Health Care System, Temple, TX
Temple, TX
Funding Period: April 2010 - March 2013

BACKGROUND/RATIONALE:
Veterans with the severe mental illnesses (SMI) of schizophrenia, bipolar disorder, post-traumatic stress disorder, and major depression (single or multiple SMIs), undergo surgery at a higher rate than veterans without SMI (2.5% vs 1.4% in fiscal 2005). These mental illnesses are common in the VA and have a devastating impact on all aspects of health. Yet our systematic literature review (1966-2007) identified only 12 studies regarding perioperative clinical outcomes for SMI patients. This meager evidence base highlights an important gap in knowledge regarding surgical outcomes of patients with SMI. Most surgical quality of care studies derive from the VA's National Surgical Quality Improvement Program (VASQIP - includes NSQIP and CICSP), however, VASQIP collects no information about preoperative psychiatric diagnoses. Therefore, VASQIP cannot even begin to explore perioperative risks and outcomes for SMI patients without incorporating additional data. SMI patients may fare poorly after surgery because of later presentation of surgical disease, lifestyle factors, and multiple chronic medical conditions. Therefore we plan to measure rates of surgery and to evaluate surgery outcomes among VA patients with and without SMI.

OBJECTIVE(S):
(1) Compare surgery rates among VA patients by SMI status (schizophrenia, bipolar disorder, PTSD, MDD, no mental illness). (2) Model 30-day, 90-day, and 1-year postoperative mortality and survival by SMI status. (3) Assess 30-day, 90-day, and 1-year postoperative complications (e.g., heart attack, venous thromboembolism, ICU admission, pneumonia, respiratory failure, sepsis, wound infection) by SMI status.

METHODS:
Our retrospective cohort study uses archival VA data to compare outcomes of surgery among patients with schizophrenia, bipolar disorder, post-traumatic stress disorder, or major depressive disorder to outcomes of surgery patients without prior diagnosis of any mental illness over the 5-year period FY05-FY09. Incorporation of VASQIP data will permit in-depth examination of risk-adjusted models and post-operative complications uniquely collected by NSQIP for the overlap sample in both data sources. We are calculating rates of surgery, postoperative mortality, postoperative survival, and postoperative complications, contrasting VA patients with and without SMI. Multivariable analyses will assess the impact of SMI on surgery outcomes, after adjusting for demographic and clinical covariates. Exploratory analyses will assess potential differences in complexity of surgical disease by SMI status using Relative Value Units (RVUs) for selected procedures, and the impact of organizational characteristics on postoperative outcomes.

FINDINGS/RESULTS:
Preliminary: We administratively combined the pilot study, POSSE (Zeber-PI) into STOPP, resulting in a cohort with rolling enrollment from FY05-FY09 (extended by 1 year into the past). The surgery cohort, using the definitions developed in the FY2005 pilot study that preceded this IIR and augmented by new codes recorded in FY2006-2009 data, now comprises 428,056 individuals including 4% DOSA surgeries, 90% inpatient surgeries, and 6% with both inpatient and DOSA procedures on the same date. These proportions reflect each patient's first qualifying surgery during the study period. Subsequent dates of surgery occur for some patients (up to 17 operations in 1 year). The cohort of surgery patients was 4% female, 18% black, 78% white, 3% other race, and 6% of patients were Hispanic. Among surgery patients, 39% of patients were VA-eligible via low income (VA Priority 1) compared to 31% of VA enrolled patients. Comorbidity status was assessed as 2.6 (SD 2.5) conditions per Charlson score, 4.3 (SD 2.3) chronic conditions per Selim score. Coronary Artery Bypass Graft (CABG) was performed on 27,753 patients, endarterectomy (ENDA) on 14,069, and arthroscopic procedures on 8,436. Roughly 9,699 of the 90,000 schizophrenia patients in the VA underwent 2,000 invasive surgeries each year. Patients with schizophrenia were less likely to experience any surgery (2.2% vs 4.1% for patients without severe mental illness (SMI)) or to receive the common surgeries (e.g. in 2006, CABG: 4% of schizophrenia vs 6% other SMI vs 7% non-SMI patients; ENDA: 3% schizophrenia vs 5% other SMI vs 6% non-SMI). FIRST MS: Among 113,726 depressed patients, 7,373 patients (6.5%) experienced one or more of the four inpatient surgeries. Patients included 7% Hispanic, 17% African American, and 2% Asian/other races. Surgeries comprised 1,441 hip/knee, 1,550 digestive, 641 CABG, and 4,311 other vascular procedures. Depressed Hispanic patients were less likely to receive hip/knee (OR=0.7), CABG (OR=0.3) or other vascular surgeries (OR=0.7) but equally likely as white or black patients to receive surgeries of the digestive system. Black patients had modestly decreased relative odds of other vascular surgery. No differences in postoperative mortality for these common inpatient surgeries were found. Readmission within 30 days of discharge following surgery, among patients with postoperative stays not exceeding 30 days, varied by SMI status. After several types of operations, patients with schizophrenia had readmission rates that were substantially higher than rates among non-SMI patients, for example, vascular operations (17% vs 22%), CABG (14% vs 24%), hip/knee procedures (8% vs 15%: nearly double).

IMPACT:
This study will produce new knowledge in the area of surgical treatment and severe mental illness. The project will produce previously unexamined data on surgical care for the population of veterans with severe mental illness, a patient subgroup with complex healthcare needs. The proposed work will lay the groundwork for identifying and developing interventions to improve perioperative care of these patients. In our first year, we found that patients with Severe Mental Illness (SMI) were over-represented among surgery patients. Although 6-7% of VA patients were classified with SMI during FY2006-FY2009, fully 17-19% of the surgery patients during the same period had SMI (proportion with SMI increased over the study period). Surgery patients were also twice as likely to be Priority 1 (23% vs 10%), three times more likely to be Priority 4 (catastrophically disabled; 8% vs 2.75%), and somewhat more likely to be Priority 5 (very low income; 39% vs 31%) relative to all patients enrolled in VA. Thus surgical teams are dealing considerably disadvantaged and mentally distressed patients. In our first submitted manuscript, we found that depressed Hispanic patients were much less likely to receive CABG surgeries compared to white or black patients (OR=0.3) although no differences in postoperative mortality was found. In our second submitted manuscript, we noted differences in 30-day readmission rates by SMI status.

PUBLICATIONS:

Journal Articles

1. Copeland LA, Zeber JE, Pugh MJ, Phillips KL, Lawrence VA. Ethnicity and Race Variations in Receipt of Surgery among Veterans with and without Depression. Depression research and treatment. 2011 Nov 10; 2011:370962.

Conference Presentations

1. Copeland LA, Zeber JE, Sako EY, Flynn J, MacCarthy A, MacCarthy D, Lawrence VA. Epidemiology of Three Common Surgeries in Veterans with Schizophrenia. Poster session presented at: International Congress on Schizophrenia Research; 2011 Apr 3; Colorado Springs, CO.
2. Zeber JE, Copeland LA, Pugh MJ, Mortensen EM, Restrepo MI, Lawrence VA. Surgical Rates of Common Procedures in Veterans with Serious Mental Illness. Poster session presented at: VA HSR&D Field-Based Mental Health and Substance Use Disorders Meeting; 2010 Apr 28; Little Rock, AR.

Journal Other

1. Zeber JE, Copeland LA, Lawrence VA, Pugh MJ, Restrepo MI. Surgical Rates of common procedures in Veterans with Schizophrenia and Other Mental Illness. [Abstract]. Schizophrenia Bulletin. 2011 Mar 1; 37(S1):101-101.

DRA: Mental, Cognitive and Behavioral Disorders
DRE: Prognosis
Keywords: Bipolar Disorder, Depression, PTSD, Schizophrenia, Serious Mental Illness
MeSH Terms: none