C. G. Ratcliff1,4,5, N. N. Massarweh2,5, S. Sansgiry5,6, L. Dindo1,5, H. Yu5,6, D. H. Berger2,5,7, J. A. Cully1,5 1Baylor College Of Medicine,Department Of Psychiatry & Behavioral Sciences,Houston, TX, USA 2Baylor College Of Medicine,Department Of Surgery,Houston, TX, USA 4Sam Houston State University,Department Of Psychology,Hunstville, TX, USA 5Michael E. DeBakey Veterans Affairs Medical Center,Houston, TX, USA 6Baylor College Of Medicine,Department Of Medicine,Houston, TX, USA 7Baylor St. Luke’s Medical Center,Houston, TX, USA
Introduction: Data regarding the impact of mental health (MH) diagnosis and treatment on postoperative outcomes are evolving. Presently, little is known about the prevalence and effect of MH treatment on outcomes following surgery for colorectal cancer (CRC).
Methods: We identified 58,961 Veterans who underwent CRC surgery from 2000-2014 using Veteran Affairs (VA) Surgical Quality Improvement Program (VASQIP) linked to the VA Corporate Data Warehouse to identify MH diagnoses and services. Multivariable logistic regression adjusting for clinical and demographic factors was used to evaluate the association between MH diagnosis (defined as depression, anxiety, PTSD, bipolar, psychotic, personality, cognitive, and substance use disorders) that were documented 30d prior to surgery and the occurrence of 1+ postoperative complication (POCOMP), 90d readmission (90dReadm), and length of stay (LOS). The impact of MH treatment (defined as psychiatric medication and psychotherapy [meds+therapy], psychiatric medication alone [meds alone], psychotherapy alone [therapy alone], or no treatment) within the 30d prior to surgery was also examined.
Results: Within the cohort, 9,029 (15%) had a MH diagnosis (depression = 2,738 [30%], anxiety = 942 [10%], PTSD = 1,762 [20%], bipolar = 505 [6%], psychotic = 679 [8%], cognitive = 239 [3%], personality = 105 [1%], substance use = 4,579 [51%]). Among Veterans with a MH diagnosis, 136 (2%) received meds+therapy, 4,157 (46%) meds alone, 308 (3%) therapy alone, and 4,428 (49%) no treatment during the 30d before surgery.
POCOMP occurred in 30% and 90dReadm in 23% of Veterans. Median LOS was 8d (IQR 6). MH diagnosis was associated with greater odds of POCOMP (OR: 1.10, CI: 1.05-1.16), 90dReadm (OR: 1.10, CI: 1.04-1.16), and longer LOS (OR: 1.42, CI: 1.09-1.86) compared to no MH diagnosis.
Veterans with a MH diagnosis who received no preoperative MH treatment (OR: 1.08, CI: 1.00-1.15) or meds alone (OR: 1.15, CI: 1.07-1.24) had greater odds of POCOMP relative to Veterans without MH diagnosis. Similarly, Veterans with a MH diagnosis who received no preoperative MH treatment (OR: 1.13, CI: 1.04-1.21) or meds alone (OR: 1.15, CI: 1.07-1.24) had greater odds of 90dReadm relative to Veterans without MH diagnosis. Finally, Veterans with a MH diagnosis who received meds alone had longer LOS relative to Veterans without MH diagnosis (OR: 1.96, CI: 1.35-2.85). Odds of POCOMP, 90dReadm, and longer LOS for Veterans with a MH diagnosis who received meds+therapy or therapy alone did not statistically differ from Veterans without MH diagnoses.
Conclusion: MH diagnoses are associated with postoperative complications and readmissions among Veterans who undergo CRC surgery. Provision of preoperative psychotherapy, alone or in combination with psychiatric medication, may help mitigate the adverse effect of psychiatric conditions. Since few Veterans receive adequate preoperative MH treatment, screening for these psychiatric risk factors may be warranted.