Apr 09, 2026
New postoperative depressive symptoms in surgical patients: a retrospective cohort analysis of a prospectively maintained cohort
- Megan Rolfzen1,
- Vidhya Gunaseelan1,
- Karsten Bartels1,
- Amy Bohnert1,2,3,4,
- Stephan Frangakis1,
- Afton Hassett1,
- Chad Brummett1,2
- 1Department of Anesthesiology, University of Michigan, Ann Arbor, MI, USA;
- 2Institute for Healthcare Policy and Innovation, University of Michigan, Ann Arbor, MI, USA;
- 3Center for Clinical Management Research, VA Ann Arbor HCS, Ann Arbor, MI, USA;
- 4Department of Psychiatry, University of Michigan Medical School, Ann Arbor, MI, USA
Protocol Citation: Megan Rolfzen, Vidhya Gunaseelan, Karsten Bartels, Amy Bohnert, Stephan Frangakis, Afton Hassett, Chad Brummett 2026. New postoperative depressive symptoms in surgical patients: a retrospective cohort analysis of a prospectively maintained cohort. protocols.io https://dx.doi.org/10.17504/protocols.io.ewov1rb9olr2/v1
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: April 08, 2026
Last Modified: April 09, 2026
Protocol Integer ID: 314669
Keywords: depression, perioperative, surgery, mental health, behavioral health, mental illness, multivariable regression, procedural risk, new postoperative depressive symptoms in surgical patient, identified postoperative depressive symptom, new postoperative depressive symptom, postoperative depressive symptom, onset depression after surgery, year prevalence rates of major depressive disorder, major depressive disorder, postoperative mortality, day postoperative mortality, onset depression, month among surgical patient, disability after surgery, depression, reported outcome measure, perioperative period due to inadequate detection method, surgical patient, mental health services administration, mental health, prolonged hospitalization, perioperative period
Funders Acknowledgements:
Society of Cardiovascular Anesthesiologists
Grant ID: In-Training Grant
National Institutes of Health
Grant ID: T90DE034663
Abstract
The most recent national prevalence study conducted by Substance Abuse and Mental Health Services Administration (SAMHSA) and Research Triangle Institute (RTI) International reports 1-year prevalence rates of major depressive disorder at 15.5%. Depression is associated with 23% increased odds of 30-day postoperative mortality, prolonged hospitalization, and disability after surgery. However, new-onset depression after surgery likely remains underdiagnosed in the perioperative period due to inadequate detection methods that rely on discrete electronic health record fields and administrative codes intended for billing purposes. This study aims to characterize newly identified postoperative depressive symptoms at one month among surgical patients by using validated patient-reported outcome measures.
Brief rationale and Hypothesis
The largest increases in age-standardised disability-adjusted life years between 2010 and 2023 stemmed from anxiety and depressive disorders.1 In the US, 80% of non-metropolitan counties do not have a single psychiatrist, roughly 80% of antidepressants are prescribed by primary care providers, and payors allocate little more than 4% of payments to therapies for mental illness and substance use addictions.2 Hospitalizations in general, and surgical episodes in particular, have been successfully leveraged to motivate tobacco cessation3 and alcohol reduction.4 The perioperative period may provide a motivating platform for functional, physical, and emotional change. With 1 in 9 adults undergoing surgery in 2018,5 leveraging surgery as the entry point for mental healthcare is a viable option.
Depression is associated with perioperative morbidity,6 a 23% increased odds of in-hospital 30-day mortality after surgery,7 and an average postoperative length of stay 1 day longer than those without preoperative depression.8 Despite the clear risks and high prevalence, perioperative depressive symptoms are underdetected due to incomplete screening penetrance and reliance on codes more relevant to surgical acute care encounters.9,10
Further research is needed to evaluate new postoperative depressive symptoms. This study aims to characterize newly identified postoperative depressive symptoms at one month among surgical patients without preoperative depression. We hypothesize that 1) current depression assessments miss a significant proportion of patients with new onset depressive symptoms and 2) individual patient and surgical characteristics are associated with de novo depressive symptoms in the postsurgical recovery period at 30 days.
Study Design
Retrospective analysis of a prospective observational cohort study.
Data Sources
Analgesic Outcomes Study, Michigan Genomics Initiative, and oPIOIDS study datasets. These prospective registries are collected and maintained at the University of Michigan. They contain linked institutional electronic health record (EHR) and patient-reported outcomes (PROs).
Inclusion Criteria
Adults >=18 years old
Patients undergoing an elective, non-emergent surgery between 2010 and 2023
Depression score available at day 1 and at month 1
Exclusion Criteria
In-hospital mortality
Preoperative depression (PROMIS depression short form 4a ≥ 8)
Explanatory Variables
Age (continuous)
Sex (categorical - F/M/other)
Ethnicity (categorical)
Race (categorical)
Relationship Status (categorical)
Education (categorical)
Surgery Type (categorical - total hip arthroplasty, total knee arthroplasty, hysterectomy, thoracic, abdominal, ankle, breast, hand, and inguinal hernia)
Physical Function (PROMIS; continuous)
Sleep Disturbance (PROMIS; continuous)
Nociplastic pain (Fibromyalgia Survey Score)
Pain Intensity (Average pain intensity; continuous)
Surgical Site Pain Intensity (Average intensity; continuous)
Anxiety (PROMIS; continuous)
Preoperative Opioid Use (Y/N)
Tobacco Use (Y/N)
Benzodiazepine Use (Y/N)
Cannabis Use (Y/N)
Charlson Comorbidity Index (continuous; 0-37)
Primary Outcome
Incidence of new postoperative depressive symptoms at one month postoperatively.
Definition: Among patients who undergo surgery, new postoperative depressive symptoms at one month with PROMIS depression short form ≥ 8 raw score among those with PROMIS depression score <8 preoperatively.
Sensitivity analysis: Alternative cohort definition. New postoperative depressive symptoms at one month is defined as PROMIS depression short form 4a ≥ 8 on the 4-item scale at 1 month postoperatively among those without an ICD diagnosis code for depression before surgery.
Secondary Outcomes
1. Depression severity (continuous measure; raw score)
2. ICD-10 depression diagnosis within 1 year post-surgery
Definition: Among patients who undergo surgery without depression preoperatively, postoperative depressive symptoms at 4 weeks will be treated as continuous (raw score; 0-20) in the primary analysis per the analytic plan below. Additionally, ICD-10 codes (F30.x-F39.x; see appendix for full list) within 1 year post-procedure will be obtained from the EHR to assess the prevalence of persistent depression based on diagnosis codes.
Statistical Analysis
**Descriptive Statistics: We will characterize the cohort's baseline characteristics and patient-reported scales using median and interquartile range for nonparametric distributions, and percentages for categorical variables.
**Primary Outcome Analysis: The primary analysis is descriptive, identifying the incidence of new depressive symptoms. Dependent on sample size, multivariable regression models will be used to identify factors associated with new-onset depressive symptoms. To evaluate an alternative cohort definition based on diagnosis codes, we will perform the primary analysis on a subgroup restricted to those without preoperative depression, identified as the absence of ICD codes for depression.
**Secondary Outcome Analysis: We will repeat the primary multivariable analysis in a generalized linear model (depending on the distribution) with depression severity scores treated as continuous. Secondarily, we will identify the percentage of those with depression based on a depression ICD-10 code within 1 year (via EHR data) post surgery date.
Limitations
1. Retrospective study
2. Data from a single state
3. Additional confounding variables
4. Cannot assume causal relationship
Protocol references
1. Hay SI, Omg KL, Santomauro DF, Bhomadewi, Aaliopur MA, Aalruz H, Aabnneh HS, Abaragou UO, Abate BB, Abafati C, Abbas N, Abbasifard M, Abbasi-Kangevari M, Abd ElHafeez S, Abdalla AN, Abdalla MA, Abdallah EM, Abebo ZA, Abdel Razek NMI, Abdelgail AM, Abdel-Hameed R, Abdelmasseh M, Abdelmalik M, Abdel-Rahman WM, Abd-Elsalam S, Abdi S, Abdollahi M, Abdoum M, Abdous A, Abdul Aziz JM, et al. Burden of 375 diseases and injuries, risk-attributable burden of 48 risk factors, and healthy life expectancy in 204 countries and territories, including 660 subnational locations, 1990–2023: a systematic analysis for the Global Burden of Disease Study 2023. Lancet 2025. Available at: http://dx.doi.org/10.1016/s0140-6736(25)01637-x.
2. How do individuals with behavioral health conditions contribute to physical and total healthcare spending? 2020. Available at: https://www.milliman.com/en/insight/how-do-individuals-with-behavioral-health-conditions-contribute-to-physical. Accessed October 19, 2025.
3. Asberg K, Bendtsen M. Perioperative digital behaviour change interventions for reducing alcohol consumption, improving dietary intake, increasing physical activity and smoking cessation: a scoping review. Perioper Med (Lond) 2021;10:118. Available at: http://dx.doi.org/10.1186/s13741-021-00189-1.
4. Egholm JW, Pedersen B, Møller AM, Adamij J, Juhl CB, Tønnesen H. Perioperative alcohol cessation intervention for postoperative complications. Cochrane Database Syst Rev 2018;11:CD008343. Available at: http://dx.doi.org/10.1002/14651858.CD008343.pub3.
5. Bicket MC, Chua K-P, Lagisetty PA, Li Y, Waljee JF, Brummett CM, Nguyen TD. Prevalence of surgery among individuals in the United States, 2024:5:e421. Available at: http://dx.doi.org/10.1097/AS9.0000000000000421.
6. Ghoniem MM, O’Hara MW. Depression and postoperative complications: an overview. BMC Surg 2016;16:5. Available at: http://dx.doi.org/10.1186/s12893-016-0120-y.
7. Abrams TE, Vaughan-Sarrazin M, Rosenthal GE. Influence of psychiatric comorbidity on surgical mortality. Arch Surg 2010;145:947–53. Available at: http://dx.doi.org/10.1001/archsurg.2010.199.
8. Panayi AC, Afxendiou A, Pantel K, Lee J, Price J, Zadeh J, Precedo A, Xiao M, Heller JA, Lakhkar B, Wiyajendraun DR, Diep C. The association between preoperative depression and length of stay after major surgery: a systematic review and meta-analysis. Anesthesiology:10.1097/ALN.0000000000005626. Available at: http://dx.doi.org/10.1097/ALN.0000000000005626.
9. Hietamies TM, Smith AE, Lii TR, Muzzall E, Flohr J, Okada RL, Andriella ZG, Nyongesa CA, Cianfichi LJ, Hack LM, Heifets BD. Prevalence and characteristics of preoperative patients with depression. Br J Anaesth 2024;133:77–85. Available at: http://dx.doi.org/10.1016/j.bja.2024.04.017.
10. Perkins ER, Latzman RD, Patrick CJ. Interfacing neural constructs with the Hierarchical Taxonomy of Psychopathology: “Why” and “how.” Personal Ment Health 2020;14:106–22. Available at: http://dx.doi.org/10.1002/pmh.1460. Accessed January 28, 2026.
11. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005;43:1130–9. Available at: http://dx.doi.org/10.1097/01.mlr.0000182534.19832.83.
12. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–9. Available at: http://dx.doi.org/10.1016/0895-4356(92)90133-8. Accessed November 1, 2025.