Methods

We conducted a nationwide, population-based, cohort study. The study was approved by the regional Human Research Ethics Committee, Stockholm, Sweden.

We identified all patients who underwent CABG in Sweden from 1997 to 2008 from the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) registry. We excluded patients who had undergone previous cardiac surgery and patients who had concomitant procedures in addition to CABG or underwent emergency surgery.

The Swedish personal identity number was used by the National Board of Health and Welfare to retrieve information from the national registries and assemble the study database. Baseline patient characteristics were obtained from the SWEDEHEART and Swedish National Patient Register (Swedish National Board of Health and Welfare). The National Patient Register covers all diagnoses for all patients hospitalized in Sweden from 1987. We used the National Patient Register (Swedish National Board of Health and Welfare) to identify patients with a diagnosis of depression ( International Classification of Disease , version 10 [ICD-10], code: F3 and ICD-9 codes: 296, 311, 300.4, 309.0, and 309.1). Patients were divided into an exposed group consisting of patients with a diagnosis of depression preceding the date of CABG surgery and an unexposed group without a preoperative diagnosis of depression.

The primary outcome measure was all-cause mortality. Secondary outcome measures included a combined end point of all-cause mortality or rehospitalization for myocardial infarction, heart failure, or stroke. Survival status was ascertained in February 2011 using the Swedish personal identity number and the continuously updated Total Population Register at Statistics Sweden. All patients with a postoperative diagnosis of myocardial infarction (ICD-10 code I21), heart failure (ICD-10 code I50), or stroke (ICD-10 codes I60 to I69) were identified. The validity of these diagnoses in the Swedish National Patient Register has been evaluated and was found to be 95% for a primary diagnosis of heart failure, and the positive predictive value for stroke was 98.6% and for myocardial infarction was 98% to 100%. Using the personal identity number and linkage to the National Patient Register, the number of days between CABG and first rehospitalization for myocardial infarction, heart failure, or stroke could be calculated. Follow-up regarding myocardial infarction, heart failure, and stroke ended on December 31, 2008. Information regarding the cause of death was obtained from the Swedish Cause of Death Register (Swedish National Board of Health and Welfare).

Time to event was calculated as the time in days from the date of CABG to the corresponding clinical end point (death from any cause or the composite end point). We used the Kaplan-Meier method to calculate cumulative survival and construct survival curves for the exposed and unexposed group and used the log-rank test to compare differences between the curves. We used Cox proportional hazards regression to model survival. We created several multivariate models considering all baseline characteristics and reached a final parsimonious model that included the following variables: age, gender, diabetes mellitus, chronic obstructive pulmonary disease, peripheral vascular disease, preoperative left ventricular function, estimated glomerular filtration rate, history of myocardial infarction, history of heart failure, history of stroke, and acute perioperative kidney injury. We checked the assumption of proportional hazards, first by graphical examination of the survival function and then by including time-dependent covariates in the final model. Model fit was evaluated by analysis of Cox-Snell residuals. Some data were missing: left ventricular ejection fraction (32%), diabetes (36%), peripheral vascular disease (32%), estimated glomerular filtration rate (17%), and acute kidney injury (36%), so multiple imputation by chained equations was used to impute missing values. The event indicator and Nelson-Aalen estimator of the cumulative baseline hazard were included in the imputation model. All analyses were performed on the imputed data set. We assumed that the missing values were missing at random. One hundred data sets were imputed, and estimates from these data sets were combined using standard methods. The objective of imputation was to retain statistical power and reduce selection bias that may occur when deleting observations with missing covariates. In addition, we performed a complete-case analysis, in which only observations without missing values for model covariates were included. Stata, version 12.1 (StataCorp LP, College Station, Texas), was used for all analyses.

Results

We identified 69,243 patients who underwent CABG from 1997 to 2008 from the SWEDEHEART register ( Figure 1 ). We excluded 1,234 patients who had previous cardiac surgery and 9,509 patients who had another cardiac procedure in addition to CABG. Finally, we excluded 2,436 patients who underwent emergency surgery defined as surgery within 24 hours of decision to operate. The final study population consisted of 56,064 patients who underwent primary, isolated, nonemergent CABG. We identified 324 patients with a preoperative diagnosis of depression from the National Patient Register. The remaining 55,740 patients were categorized as the control group. Baseline characteristics of the study population are listed in Table 1 . Patients with depression and the control group were not balanced regarding several potentially confounding factors, most importantly, female gender, diabetes mellitus, peripheral vascular disease, chronic obstructive pulmonary disease, and a history of stroke.

Flowchart of patient exclusion criteria.

The total follow-up time was 422,162 person-years (mean follow-up time 7.5 years), and there was no loss to follow-up. During follow-up, 114 patients (35.1%) with depression died, compared with 13,767 patients (24.6%) in the control group. The early mortality (death within 30 days of surgery) in patients with depression was 3.1%, and it was 1.5% in the control group. The age- and gender-adjusted relative risk for early mortality associated with depression was 2.06 (95% confidence interval [CI] 1.09 to 3.89). After multivariate adjustment, early mortality was not significantly greater in patients with depression (odds ratio 1.33, 95% CI 0.68 to 2.61).

The crude and multivariate-adjusted associations between depression and all-cause mortality are listed in Table 2 .

In the unadjusted Cox regression analysis, depression was associated with increased mortality (hazard ratio [HR] 1.94, 95% CI 1.61 to 2.33). The Kaplan-Meier estimates of survival are shown in Figure 2 . The unadjusted 1-, 5-, and 12-year survival was 93%, 80%, and 41% in depressed patients and 97%, 89%, and 63% in the control group (p <0.001), respectively. In the final multivariate Cox regression model, depression was significantly associated with increased mortality (HR 1.65, 95% CI 1.37 to 1.99). Depression was more common among women compared with men (1.1% vs 0.43%). The crude and multivariate-adjusted associations between depression and all-cause mortality stratified by gender are listed in Table 3 . There was a significant association between depression and mortality in men (HR 1.96, 95% CI 1.55 to 2.47) but not in women (HR 1.24, 95% CI 0.90 to 1.71).

Kaplan-Meier estimated survival in 56,064 patients with or without depression diagnosis before primary isolated CABG from 1997 to 2008 in Sweden.

Stay updated, free articles. Join our Telegram channel

Join