Impact of post-sepsis cardiovascular complications on mortality in sepsis survivors: a population-based study

Background It remains unclear whether sepsis-related cardiovascular complications have an adverse impact on survival independent of pre-existing comorbidities. To investigate the survival impact of post-sepsis cardiovascular complications among sepsis survivors, we conducted a population-based study using the National Health Insurance Database of Taiwan. Methods We identified sepsis patients from the National Health Insurance Research Database of Taiwan using ICD-9-CM codes involving infection and organ dysfunction between 2000 and 2011. Post-sepsis incident myocardial infarction (MI) and stroke were ascertained by ICD-9-CM codes and antiplatelet treatment. We constructed a non-sepsis comparison cohort using propensity score matching to ascertain the association between sepsis and cardiovascular complications. Furthermore, we compared the 180-day mortality and 365-day mortality between patients surviving sepsis with or without post-sepsis MI or stroke within 70 days of hospital discharge. We constructed Cox regression models adjusting for pre-existing comorbidities to evaluate the independent survival impact of post-sepsis MI or stroke among sepsis survivors. Results We identified 42,316 patients hospitalized for sepsis, from which we matched 42,151 patients 1:1 with 42,151 patients hospitalized without sepsis. Compared to patients hospitalized without sepsis, patients hospitalized with sepsis had an increased risk of MI or stroke (adjusted odds ratio 1.72, 95% CI 1.60–1.85). Among 42,316 patients hospitalized for sepsis, 486 (1.15%) patients developed incident stroke and 108 (0.26%) developed incident MI within 70 days of hospital discharge. Compared to sepsis survivors without cardiovascular complications, sepsis survivors with incident MI or stroke had a higher mortality rate at 180 days (11.68% vs. 4.44%, P = 0.003) and at 365 days (16.75% vs. 7.11%, P = 0.005). Adjusting for age, sex, and comorbidities, post-sepsis MI or stroke was independently associated with increased 180-day (adjusted hazard ratio [HR] 2.16, 95% CI 1.69–2.76) and 365-day (adjusted HR 1.90, 95% CI 1.54–2.32) mortality. Conclusions Compared to sepsis patients without incident MI or stroke, sepsis patients with incident MI or stroke following hospital discharge had an increased risk of mortality for up to 365 days of follow-up. This increased risk cannot be explained by pre-sepsis comorbidities. Electronic supplementary material The online version of this article (10.1186/s13054-019-2579-2) contains supplementary material, which is available to authorized users.


Background
Sepsis, which refers to life-threatening organ dysfunction caused by a dysregulated host response to infection, is a major public health concern [1]. In Taiwan, the incidence of sepsis has steadily risen from 638 per 100,000 persons to 772 per 100,000 persons from 2002 to 2012, while the annual mortality has steadily decreased from 27.8 to 22.8% over the same time period [2]. While short-term sepsis mortality is decreasing, mid-to longterm sepsis mortality has remained high as many patients die in the subsequent months. Recent studies have suggested that the increased risk of mid-to long-term mortality after sepsis cannot be explained by the preexisting comorbid conditions before sepsis and might be attributed to increased post-sepsis cardiovascular complications instead [3]. Sepsis increases the risks of cardiovascular complications during and shortly after admission, with up to fourfold increase in the risk [4][5][6][7]. The increased risk of cardiovascular complications has been attributed to a variety of pathophysiologic mechanisms, including immunoparalysis, depression of ventricular function, arrhythmia, organ ischemia related to increased oxygen demand, procoagulant changes in the blood, impaired cardiovascular autonomic response, and accelerated atherosclerosis [8][9][10]. Consistent with prior literature, our previous work using the National Health Insurance Research Database of Taiwan demonstrated that patients with sepsis are at markedly elevated risk of incident MI/stroke during the first 70 days after hospital discharge [11]. We found that after this critical 70-day period, sepsis survivors have a comparable risk of post-sepsis MI/stroke with non-sepsis control patients.
Despite the established association between sepsis and cardiovascular complications, limited information is known about the survival impact of these post-sepsis cardiovascular complications. Among the difficulties in assessing the survival impact of post-sepsis cardiovascular complications are the rarity of post-sepsis cardiovascular complications and the complex competing risk relationship between mortality and post-sepsis incident cardiovascular complications [12,13]. The aim of this study was twofold. First, we sought to determine the independent association of sepsis with incident cardiovascular disease after sepsis hospitalization with the construction of a non-sepsis comparison group. Second, we sought to ascertain if the increase in mortality associated with cardiovascular events in sepsis patients exceeded the mortality associated with cardiovascular events in non-sepsis patients. The use of a national population-based database allows us to build a large sepsis cohort with a sufficient number of events to compare the outcome between patients with and without incident MI/stroke after sepsis [14,15].

Data source
The present study used the National Health Insurance Research Database (NHIRD) of Taiwan. Taiwan's National Health Insurance (NHI) program is a single-payer mandatory health insurance system that covers over 99% of the 23,000,000 people residing in Taiwan. The database used a systematic sampling strategy to select one million participants who were representative of the demographic and geographic region distribution of Taiwan in 2001. This sample, hereafter referred to as the one million longitudinal sample, was followed from 2001 to 2012 to form a closed cohort for research purposes. The database collects outpatient and inpatient electronic records on demographics, eligibility, vital status, diagnoses (International Classification of Diseases, ninth revision, Clinical Modification [ICD-9-CM]), operations, and prescriptions. All claims can be linked in chronological order to provide a temporal sequence of all health services utilization. Patient consent was not required for this study as this was an anonymized electronic database study. This study was approved by the Institutional Review Board of National Taiwan University Hospital. This observational study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational studies [16].

Design and study participants
The study was a population-based cohort study, consisting of all emergency department (ED) or hospital-treated sepsis patients between 2001 and 2011.

Identification of sepsis cases
Sepsis patients were identified based on the coding system used by Angus et al., which is considered to most closely resemble the latest Sepsis-3 definition among the published algorithms [17]. Operationally, sepsis patients were identified using ICD-9-CM codes for the presence of either a bacterial or fungal infection (Additional file 4: Appendix 1) plus dysfunction of one or more organ systems (Additional file 4: Appendix 2) [1]. The ICD-9-CM codes for the identification of infections in this study were the same as those used by Angus et al. (including 1286 distinct infection codes). The ICD-9-CM codes for acute organ dysfunctions were supplemented with NHI procedure codes to increase specificity. In this manuscript, MI refers to myocardial infarction alone, stroke refers to cerebrovascular events alone, and MI/stroke stands for the combined outcome of MI and stroke.

Construction of the non-sepsis comparison cohort
To ascertain the independent association between sepsis and incident myocardial infarction (MI)/stroke after sepsis hospitalization, we constructed a non-sepsis comparison cohort using the risk set sampling and propensity score matching technique. The non-sepsis comparison cohort was constructed using a two-stage procedure. In the first stage, we used the risk set sampling method to sample 100 non-sepsis patients for each sepsis case, matching on admission date, 5-year age group, sex, and quartile Charlson Comorbidity Score (0, 1-2, 3-4 and ≥ 5) [18]. In the second stage, we created a propensity score (PS) consisting of a comprehensive set of covariates associated with sepsis (Additional file 1: Table S1). We performed 1:1 PS matching using a greedy algorithm to construct the final non-sepsis comparison cohort. The cohort construction process is shown in Fig. 1.

Identification of post-sepsis cardiovascular complications
As the primary aim was to analyze the impact of postsepsis stroke/MI on mid-or long-term survival, the primary "exposure" in this study was incident MI/stroke after sepsis hospitalization. Our previous study found that the critical period for MI/stroke is 70 days after discharge from a sepsis hospitalization. We therefore assessed for MI/stroke within the 70-day risk period as the primary exposure of this study. Patients with stroke were identified by the presence of an ICD-9-CM diagnosis code of 433.xx or 434.xx, which have a positive predictive value (PPV) of 0.96 and 1.00, respectively, in the NHIRD database [19]. Patients with incident acute MI were identified by any primary or secondary admission diagnosis containing ICD-9-CM code of 410.xx together with a prescription for antiplatelet therapy, such as aspirin, clopidogrel, dipyridamole, and ticlopidine. Newer classes of antiplatelets were not available in Taiwan during the study period. The above search algorithm yields a PPV of 0.84 for acute MI in the NHIRD database [20]. The primary outcomes of interest were 180-day and 365-day all-cause mortality.

Follow-up of cohort
For both the primary sepsis cohort and non-sepsis comparison cohort, we anchored the 71st day after hospital discharge as the index day of cohort entry. Patients were followed up for three outcomes: death, termination of health insurance coverage, and end of study (the 365th day from the index date), whichever came first. To control for unbalanced covariates between patients with/ without post-sepsis cardiovascular events, we collected information on demographics, urbanization level, insurance premium level, chronic comorbidities, and risk factors for sepsis. In order to remove the effect of covariates that could develop after incident cardiovascular events, we collected all covariate information from the beginning of the study (year 2001) to the discharge date of the index hospitalization. The level of urbanization of the cities/towns was stratified into four levels based upon a composite score obtained by calculating population density (people/km 2 ), population percent of people with an educational level of college or above (%), percent of people over 65 years (%), percent of agriculture workers (%), and the number of physicians per 100,000 people. To further control for general health conditions that are not reflected by ICD-9 CM codes [21], we further collected the frequency of healthcare facility utilization in the 1-year period before sepsis admission as a proxy indicator of general health. The timeline of the study design and periods of data collection can be seen in Fig. 2.

Data analysis
The analysis was implemented in two stages. In the primary analysis, we analyzed the impact of incident MI/ stroke on the mid-or long-term outcome of sepsis survivors. In the secondary analysis, we compared the incidence and outcome of incident stoke/MI after hospital discharge between the sepsis and non-sepsis cohorts. In the primary analysis, we first compared the baseline characteristics between sepsis survivors who developed incident acute MI/stroke and those who did not develop incident acute MI/stroke. Categorical variables were presented as frequency and percentage and compared using chi-squared tests. Continuous variables were presented as mean or median and compared by non-parametric Mann-Whitney U tests. We then plotted the cumulative mortality curves of patients with/without post-sepsis MI/stroke. The differences in cumulative mortality between the groups were compared using the Wilcoxon signed-rank test. The survival impact of post-sepsis MI/ stroke was assessed by multivariable Cox regression analyses. The proportional hazards assumption was checked by plotting curves of the log of the negative log of the survival function against log time for patients with/without post-sepsis MI/stroke. Potential confounders were determined based on the potential relevance given the literature, including the following variables: demographics, urbanization level, insurance premium, comorbidity, and healthcare service utilization. In the secondary analysis, we constructed a non-sepsis comparison cohort Fig. 2 Timeline for the study design. During the study period from 2001 to 2011, patients who had post-sepsis myocardial infarction or stroke within 70 days of the last hospital discharge due to sepsis entered the cohort on the 71st day. Three outcomes were followed: the end of the 365-day study period, death, or termination of health insurance coverage using the PS matching method mentioned above. To evaluate the success of the matching process, we calculated the standardized difference of matching covariates between the sepsis cohort and the non-sepsis comparison cohort. To evaluate the independent association between sepsis and post-sepsis MI/stroke, we calculated the incidence and 180-day mortality of MI/stroke among patients for the sepsis cohort and the PS-matched nonsepsis comparison cohort. We used univariate conditional logistic regression to evaluate the risk of incident MI/stroke and stratified Cox regression analysis to evaluate mortality risk. All statistical analyses were performed in SAS 9.4 (SAS Inc. Cary NC). In all analyses, a P value of < 0.05 was deemed significant.

Baseline characteristics of study patients
From one million NHIRD participants, we identified 42, 316 patients who were hospitalized with sepsis ( Fig. 1). There were 591 events observed in the 70-day risk period after discharge, of which 486 (82.2%) were stroke, 108 (18.3%) were MI, and 3 (0.5%) were concomitant stroke and MI. The preadmission characteristics of sepsis patients with/without post-sepsis MI/stroke are summarized in Table 1. Compared with patients without MI/stroke, patients who developed MI/stroke were older, associated with lower socioeconomic status, had a higher burden of preadmission comorbidities, and had a higher frequency of healthcare service utilization.

Survival impact of post-sepsis acute MI and stroke
The 180-day mortality for sepsis survivors without incident MI/stroke was 4.61% (1450/31,423  Table 3.

Comparison between sepsis and non-sepsis cohorts
Of the 42,316 identified sepsis patients, 41,251 were matched to a non-sepsis cohort by PS matching. The standardized difference of matching covariates between sepsis and non-sepsis patients was less than 6%, indicating a successful match (Additional file 1:  Table S2). Finally, we compared mortality between the two groups of patients. For the sepsis cohort (n = 41,251), the 180-day mortality for MI, stroke, and composite MI/stroke was 13%, 9%, 10%, respectively. On the other hand, for the non-sepsis cohort (n = 41,251), the 180-day mortality for MI, stroke, and composite MI/stroke was 13%, 10%, and 11%, respectively. Cox regression showed sepsis was not associated with an increased hazard of mortality after MI, stroke, or MI/stroke (Additional file 3: Table S3).

Discussion
Sepsis is a major public health problem. Mortality associated with sepsis may be due to post-sepsis complications. In this comparison of patients hospitalized with sepsis and patients hospitalized without sepsis matched by PS, we showed that the increased risk of cardiovascular events after sepsis hospitalization is due to sepsis per se rather than a more generalized effect of acute illness, hospitalization, or deterioration in health status. We also showed that patients who developed cardiovascular events after sepsis hospitalization have an approximately twofold increase in 180-day or 365-day mortality as compared with patients who did not sustain cardiovascular events after sepsis hospitalization. However, survival after post-hospitalization MI/stroke did not differ between sepsis and non-sepsis patients. Several major studies have shown that infection (e.g., pneumonia, sepsis) is associated with increased risk of  [4].
In this study, we further investigated a critical yet unresolved question: whether post-sepsis MI/stroke has an independent impact on the mid-and long-term survival among sepsis survivors. Consistent with the findings by Smilowitz et al., we showed post-sepsis MI/stroke itself is an independent risk factor for mortality among sepsis survivors. Smilowitz et al. showed that the in-hospital mortality is higher among patients who had both sepsis and MI as compared with patients who had sepsis alone (adjusted odds ratio 1.24; 95% CI 1.22-1.26) [22]. However, the aforementioned study did not address the potential survivor bias [23]. We previously found the first 70 days after hospital discharge to be a period of increased susceptibility to incident cardiovascular events for sepsis patients. In addition, the mortality rate was high in the first few weeks after discharge from a sepsis hospitalization. Therefore, we indexed the cohort entry at day 70 after hospital discharge, thereby mitigating the influence of potential survivor bias or competing risk by early death. Wang et al. analyzed the impact of MI on sepsis survival, which was based on the same hospital stay. In contrast, using population-based longitudinal follow-up data, we were able to provide the evidence on the long-term outcome in patients associated with postsepsis cardiovascular events. As shown in the current study, cardiovascular events had a major adverse impact on long-term survival of sepsis patients, suggesting a need to monitor for cardiovascular events among sepsis survivors.    [24,25]. The relatively low mortality observed in our cohort could be explained by the different index dates between the current studies and previous studies. Our study design built in a lag time between hospital discharge and index date, which prevents us from gaging the high mortality of sepsis patients within this critical 70-day window. The mortality of this cohort in the 70-day period after hospital discharge was as high as 24.3%. The exclusion of patients who died within 70 days of discharge resulted in a low mortality rate in our patients.
The findings of this study support the hypothesis that the long-term mortality of sepsis is at least partially mediated through post-sepsis MI/stroke. The observed adverse survival impact of post-sepsis MI/stroke implicates the potential benefits from early preventative measures. Among the potential mediators of post-sepsis cardiovascular complications [26][27][28][29][30][31], platelets might serve as a major target for pharmacological intervention [32][33][34][35][36]. The protective role of long-term aspirin use in the risk of cardiovascular events is well documented [37][38][39][40], justifying future trials on its preventive role in the cardiovascular events after a sepsis episode. Such a trial may also consider including sepsis survivors without traditional cardiac risk factors. Another promising class of pharmacologic preventive agents is statins, which also have a well-established established role in the prevention of cardiovascular events. Aside from lipid-lowering properties, the pleiotropic effects of statins including anti-inflammatory and recently discovered bactericidal effects may offer added benefit to sepsis survivors [41,42]. Definitive roles of these medications for sepsis patients awaits further investigation.
Our study has several strengths. The National Taiwan Health Insurance Database is large and comprehensive and has lent itself to many longitudinal analyses of sepsis. We used the Angus System for sepsis case identification, which is the best available algorithm for an administrative database. We used Cox regression and limited to cardiovascular events within 70 days of the sepsis event to mitigate concerns about survival time bias.
The results of this study should be interpreted in light of a few limitations. First, like other administrative databases, certain important risk factors for cardiovascular disease such as body mass index, smoking status, and alcohol consumption are lacking. Second, reverse causation may be at play given that MI or cerebrovascular events might both predispose patients to sepsis and impact the patients' survival. To avoid the potential reverse causation, post-sepsis MI/stroke was ascertained after the hospital discharge from sepsis to establish temporality, and multivariable analyses were used to adjust for confounding effects introduced by prior myocardial infarction or cerebrovascular events. Third, with respect to our finding that there was no association between sepsis and subsequent MI, two potential causes may underlie this phenomenon. First, the comparison cohort consisted of hospitalized patients, who were already at elevated risk of MI, rather than healthy population controls. Second, the number of patients with incident MI in the sepsis cohort and the comparison cohort was only 251 and 228, respectively. As such, we did not have sufficient statistical power to detect this subtle difference.

Conclusions
Sepsis is a major public health problem, and mortality associated with sepsis may be due to the complications that develop or accelerate after the sepsis episode. We demonstrated that the increased risk of cardiovascular events after sepsis is due to sepsis per se rather than a more generalized effect of deteriorating health status. Incident cardiovascular events after sepsis have a strong adverse impact on long-term survival. Current work may inform the need to extend sepsis care from acute stage management to the prevention of cardiovascular complications in the convalescent stage.

Additional files
Additional file 1: Table S1. Characteristics of sepsis patients and propensity score-matched non-sepsis patients. Summary and comparison of the demographics and underlying comorbidities of the sepsis cohort and the non-sepsis cohort. Abbreviations: MI, myocardial infarction; AIDS/ HIV, acquired immune deficiency syndrome/human immunodeficiency virus. (DOCX 17 kb) Additional file 2: Table S2. Comparison of incidence (events/1000 person) of MI/stroke in sepsis and propensity score-matched non-sepsis cohort. Representation of the multivariable logistic regression analysis summarizing the association between sepsis and incidence of MI, stroke and MI/stroke while accounting for potential confounders. Abbreviations: MI, myocardial infarction; PS, propensity score; OR, odds ratio. (DOCX 15 kb) Additional file 3: Table S3. Comparison of mortality of MI/stroke in sepsis and propensity score-matched non-sepsis cohorts. Representation of the survival analysis results using Cox regression model summarizing survival impact of sepsis on the mortality of MI, stroke and MI/stroke while accounting for potential confounders. Abbreviations: MI, myocardial infarction; HR, hazard ratio. (DOCX 15 kb) Additional file 4: Appendix 1. Codes associated with infection. The ICD-9-CM codes utilized to identify infections. Appendix 2. Codes associated with organ dysfunction. The ICD-9-CM codes utilized to identify organ dysfunction. (DOCX 29 kb)