Risk Factors for Long-term Mortality and Patterns of End-of-Life Care Among Medicare Sepsis Survivors Discharged to Home Health Care | Critical Care Medicine | JAMA Network Open | JAMA Network
[Skip to Navigation]
Figure 1.  Flow Diagram for Identification of Cohort of Medicare Sepsis Survivors Discharged to Home Health Care
Flow Diagram for Identification of Cohort of Medicare Sepsis Survivors Discharged to Home Health Care
Figure 2.  One-Year Survival Among 87 581 Patients After Sepsis Discharge to Home Health Care
One-Year Survival Among 87 581 Patients After Sepsis Discharge to Home Health Care
Table 1.  Characteristics of 87 581 Patients in the Medicare Sepsis Survivor Cohort Discharged to Home Health Care
Characteristics of 87 581 Patients in the Medicare Sepsis Survivor Cohort Discharged to Home Health Care
Table 2.  Acute Care and Hospice Use Among 24 423 Decedents
Acute Care and Hospice Use Among 24 423 Decedents
Table 3.  Multivariate Regression Results for 1-Year Mortality Among All Sepsis Survivors and Hospice Enrollment Among Decedents
Multivariate Regression Results for 1-Year Mortality Among All Sepsis Survivors and Hospice Enrollment Among Decedents
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Critical Care Medicine
    February 26, 2020

    Risk Factors for Long-term Mortality and Patterns of End-of-Life Care Among Medicare Sepsis Survivors Discharged to Home Health Care

    Author Affiliations
    • 1Division of Pulmonary, Allergy, and Critical Care, Perelman School of Medicine, University of Pennsylvania, Philadelphia
    • 2Palliative and Advanced Illness Research (PAIR) Center, Perelman School of Medicine, University of Pennsylvania, Philadelphia
    • 3Center for Home Care Policy & Research, Visiting Nurse Service of New York, New York
    • 4Department of Economics, Hunter College, The City University of New York (CUNY), New York
    • 5National Bureau of Economic Research, Cambridge, Massachusetts
    • 6Department of Biobehavioral Health Sciences, School of Nursing, University of Pennsylvania, Philadelphia
    JAMA Netw Open. 2020;3(2):e200038. doi:10.1001/jamanetworkopen.2020.0038
    Key Points español 中文 (chinese)

    Question  What are the risk factors for long-term mortality and patterns of end-of-life care among sepsis survivors who are Medicare beneficiaries and have been discharged to home health care?

    Findings  In this cohort study of 87 581 adult sepsis survivors who are Medicare fee-for-service beneficiaries and have been discharged to home health care, 1 in 4 survivors died within 1 year, and among the decedents, hospitalization and intensive care unit use in the last 30 days of life and in-hospital death were common. Several factors were found to be associated with an increased risk of mortality.

    Meaning  The findings of this study suggest that home health assessments may provide an opportunity to identify high-risk sepsis survivors and target efforts to improve their end-of-life care.


    Importance  Despite a growing recognition of the increased mortality risk among sepsis survivors, little is known about the patterns of end-of-life care among this population.

    Objective  To describe patterns of end-of-life care among a national sample of sepsis survivors and identify factors associated with long-term mortality risk and hospice use.

    Design, Setting, and Participants  This cohort study assessed sepsis survivors who were Medicare fee-for-service beneficiaries discharged to home health care using national Medicare administrative, claims, and home health assessment data from 2013 to 2014. The initial and final primary analyses were conducted in July 2017 and from July to August 2019, respectively.

    Exposures  Sepsis hospital discharge and 1 or more home health assessments within 1 week.

    Main Outcomes and Measures  Outcomes were 1-year mortality among all sepsis survivors and hospitalization in the last 30 days of life, death in an acute care hospital, and hospice use among decedents. Multivariate logistic regression was used to identify factors associated with 1-year mortality and hospice use.

    Results  Among 87 581 sepsis survivors who were Medicare fee-for-service beneficiaries discharged to home health care, 49 323 (56.3%) were aged 75 years or older, 69 499 (79.4%) were non-Hispanic white, and 48 472 (55.3%) were female. Among the total survivors, 24 423 (27.9%) people died within 1 year of discharge, with a median (interquartile range) survival time of 119 (51-220) days. Among these decedents, 16 684 (68.2%) were hospitalized in the last 30 days of life, 6560 (26.8%) died in an acute care hospital, and 12 573 (51.4%) were enrolled in hospice, with 5729 (45.6%) using hospice for 7 or fewer days. Several factors were associated with 1-year mortality, including a cancer diagnosis (odds ratio [OR], 3.66; 95% CI, 3.50-3.83; P < .001), multiple dependencies of activities of daily living or instrumental activities of daily living (OR, 2.80; 95% CI, 2.57-3.05; P < .001), and an overall poor health status (OR, 2.21; 95% CI, 2.01-2.44; P < .001) documented on home health assessment. Among the decedents, cancer was associated with hospice use (OR, 2.25; 95% CI, 2.11-2.41; P < .001), patients aged 85 years or older (OR, 1.49; 95% CI, 1.37-1.61; P < .001), and living in an assisted living setting (OR, 1.93; 95% CI, 1.69-2.19; P < .001).

    Conclusions and Relevance  The findings of this study suggest that death within 1 year after sepsis discharge may be common among Medicare beneficiaries discharged to home health care. Although 1 in 2 decedents used hospice, aggressive care near the end of life and late hospice referral were common. Readily identifiable risk factors suggest opportunities to target efforts to improve palliative and end-of-life care among high-risk sepsis survivors.


    Although sepsis is common, reductions in hospital mortality have led to an increasing number of survivors,1,2 with more than 1 million patients being discharged after treatment of sepsis from United States hospitals each year.3 However, sepsis survivorship often comes at a cost, namely reduced health-related quality of life, cognitive and functional impairments,4 increased risks of hospital readmission,5-7 and long-term mortality risk.8-11 Amid a heightened awareness of these long-term consequences,12,13 improving the quality of post-sepsis care has become a global priority.13

    Efforts to improve outcomes among sepsis survivors have largely focused on preventing or mitigating postdischarge morbidity and mortality. Nearly 1 in 2 sepsis survivors receive postacute care services, such as home health care, inpatient rehabilitation, and skilled nursing facility placement.5,14,15 Yet, the long-term risk of death after sepsis remains high compared with hospitalized patients without sepsis.7,9,16 Despite mounting evidence that sepsis survivorship is associated with increased mortality, little attention has been paid to the patterns of end-of-life care among this population. This is a particularly important evidence gap to fill in sepsis research considering the sustained national focus on improving the quality of end-of-life care for all seriously ill adults.17,18 Furthermore, as policy recommendations regarding optimal advance care planning practice and use of community-based palliative care services continue to evolve,19-22 it is important for sepsis survivors to be recognized as a potential population in need of these services.

    In this national study, we assessed the risk of long-term mortality and end-of-life care among Medicare beneficiaries discharged to home health care after sepsis. Homes are common postacute care destinations after sepsis, second only to skilled care facilities.5,14,15 Annually, approximately 200 000 sepsis survivors are discharged to home to receive health care services such as skilled nursing, physical and occupational therapy, and health aid visits.15 Because home health care spending is projected to outpace any other national health expenditure over the next decade,14 it is essential to evaluate outcomes among this population. In the present study, 1-year mortality rates were assessed and factors associated with mortality were identified from the sepsis hospitalization and the initial home health assessment, including functional assessments. The rates of hospitalization among the decedents in the last 30 days of life, in-hospital death, and hospice use, were assessed and characteristics associated with hospice use were identified.

    Data Sources and Study Population

    This retrospective cohort study used Medicare administrative and claims files from calendar years 2013 and 2014 to identify patients with sepsis who were hospitalized and discharged to home health care between July 1, 2013, and December 31, 2013, and to evaluate hospice and mortality outcomes up to 1 year after hospital discharge. The files used included the Medicare Beneficiary Summary file, Medicare Inpatient Standard Analytic File (SAF), Outpatient SAF, Home Health SAF, Hospice SAF, Part B SAF, Chronic Conditions Warehouse (CCW), and US Census data.52 We linked these files with the Outcome and Assessment Information Set (OASIS)–C,23,24 a comprehensive, federally mandated assessment of patients’ health, social, cognitive, and functional status, which is completed on initiation of home health care services. This study was approved by the respective institutional review boards of the Visiting Nurse Service of New York and the University of Pennsylvania, with a waiver of informed consent under category 4 of the Federal Policy for the Protection of Human Subjects (45 CFR 46) in 2015. This report followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.25 Initial and final primary analyses were conducted in July 2017 and July and August 2019, respectively.

    Sepsis was defined using a combination of 2 strategies given the limited sensitivity of sepsis identification from administrative claims.26 First, we used the International Classification of Diseases and Related Health Problems, Ninth Revision, Clinical Modification (ICD-9-CM) codes 995.91 (sepsis without organ dysfunction), 995.92 (severe sepsis), and 785.52 (septic shock), which were added to the ICD-9-CM codes in 2003 to improve accuracy of case identification.26 Second, we used the implicit approach developed by Angus and colleagues27 that requires an ICD-9-CM code for infection and end-organ dysfunction, which was initially developed using 1995 claims data and was subsequently validated using administrative claims from 2009 to 201028 and medical records from 2005 to 2009.26

    Beneficiaries had to be at least 18 years of age and have received at least 1 home health care visit within 1 week of discharge. Beneficiaries with additional health care use (hospital readmission, observation unit stay, and hospice admission) before their first home health visit and those without complete OASIS-C data were excluded from the study. Finally, we included only the index sepsis discharge to home health care in this sample to ensure independence of observations.


    The 1-year mortality among all sepsis survivors and hospice use among decedents were examined. The date of death was obtained from Medicare records and time to death was calculated from the discharge date of index hospital stay. The Inpatient SAF was used to identify hospitalization and intensive care unit (ICU) use within the last 30 days of life and acute care hospital as site of death. Hospice enrollment and length of stay were identified using the Hospice SAF, with late hospice referral defined as a hospice admission date 7 or fewer days prior to the date of death.29-31

    Individual Characteristics

    Medicare administrative data provided patients’ demographic characteristics, comorbidities, and clinical characteristics from the index sepsis discharge, including the admission type (Medicare Severity–Diagnosis Related Group [MS-DRG]), ICU use, and infection source. To address underreporting in claims data, ethnicity and median family income in the county where the patient lived were obtained from OASIS-C and census data, respectively, and a diagnosis of Alzheimer disease and associated dementias was obtained from the CCW.

    The initial OASIS-C assessment was conducted by a trained home health clinician at the start of a new episode of home health care within 2 days of hospital discharge for 81% of the cohort (n = 70 941) and within 7 days for the remaining sample (n = 16 640). More than 100 items were assessed, including activities of daily living (ADLs), instrumental activities of daily living (IADLs), living arrangements, cognitive functioning, sensory and behavioral status, disease signs and symptoms by organ system, frailty, and overall health status.32,33

    Statistical Analyses

    Variables were summarized using frequencies and proportions for categorical data or means (SDs) and medians (interquartile ranges [IQRs]) for continuous data. Bivariate analyses were performed using a χ2 test to compare patient characteristics between decedents and survivors. Survival data were expressed as medians (IQRs), and analyses were performed using the Kaplan-Meier method, with censoring of all patients who remained alive 365 days after sepsis discharge.

    Two multivariate logistic regression models were built using forward selection to examine the independent associations between patient characteristics and 1-year mortality among all sepsis survivors and hospice use among the decedents. Candidate covariates were selected a priori based on existing literature and clinical expertise and included those variables with a 2-sided P < .05 in the final multivariate models. Variance inflation factor diagnostic tests were used to check for collinearity between covariates and those with a value greater than 10 were excluded from the final models.34 Both final models included the following characteristics known before and during the index sepsis hospitalization: age, race/ethnicity, Medicaid status, comorbidities,35 sepsis severity, hospital-acquired sepsis, infection source, ICU admission, and surgical admission type. Items from the postdischarge OASIS-C home health assessment included in the final models are provided in Table 1 and include the following: risk for hospitalization, overall health status, living arrangement, impaired vision, number of medications, dyspnea, cognitive impairment, and the number of ADL or IADL dependencies. Categorical variables were collapsed in the final models based on the distribution of responses. Data completeness was excellent; 168 patients (1.9%) with an unknown or unclear or missing response for the overall status item on the home health assessment were excluded from the models.

    In secondary analyses, we included only covariates with a bivariate association of P < .001 to avoid overfitting the models. The results of these parsimonious models remained unchanged; thus, the results of the full explanatory models are presented. Analyses stratified by the presence of a cancer diagnosis, which has previously been shown to be associated with in-hospital mortality and hospice referral among sepsis readmissions, were also performed.8

    For all analyses, given the large sample size, statistical testing was 2-sided with a significance threshold of P < .001, and results were preferentially judged by their clinical relevance. RStudio, version 3.3.1 (R Core Team), was used for descriptive statistics and statistical analyses. SAS, version 9.4 (SAS Institute Inc) was used for diagnostic variance inflation factor tests and Kaplan-Meier plot generation.


    The cohort included 87 581 sepsis survivors with Medicare insurance who were discharged to new home health care services between July 1, 2013, and December 31, 2013 (Figure 1). Among them, 49 323 (56.3%) patients were aged 75 years or older, 69 499 (79.4%) were non-Hispanic white, and 48 472 (55.3%) were women. Overall, patients had a mean (SD) total number of comorbidities of 4.35 (1.89), and 64 200 (73.3%) were not Medicaid-eligible (Table 1). Severe sepsis was observed in most cases (70 513 [80.5%]), with the genitourinary system being the most common infection source (36 202 [41.3%]), and only half of the patients (44 241 [50.5%]) having received care in an ICU during the index sepsis hospitalization. Home health care assessments after discharge suggested that half of the patients had at least 1 indicator of risk for hospitalization (43 993 [50.2%]) and dependency in more than 2 ADLs/IADLs (76 871 [87.8%]). Uncontrolled symptoms were common, with 46 512 (51.3%) of patients experiencing pain daily or constantly and 68 525 (78.3%) having dyspnea on exertion or at rest. Nearly half of the patients had at least mild limitations in cognitive function (41 532 [47.3%]) and speech (36 751 [41.9%]).

    Nursing visits, which were received by 83 537 [95.4%] patients, were the most common home health care service provided within 7 days of sepsis discharge, followed by physical therapy (46 096 [52.6%]), occupational therapy (14 114 [16.1%]), and speech therapy (1897 [2.2%]). During the first week, 34 064 (38.9%) sepsis survivors were seen in the ambulatory setting by a medical professional.

    Mortality and End-of-Life Health Care Use

    Among the total survivors, 24 423 (27.9%) patients died within 1 year of discharge, with a median (IQR) survival time of 119 (51-220) days (Figure 2). The characteristics of decedents and survivors at 1 year after discharge differed in several patient-related, sepsis-related, and home health assessment characteristics (eTables 1 and 2 in the Supplement), most notably in the proportion of patients with a cancer diagnosis (decedents, 5965 [24.4%] vs survivors, 5462 [8.2%]; P < .001). Among all decedents, 16 684 (68.2%) were hospitalized, 10 190 (61.1%) were admitted to an ICU during the last 30 days of life, and 6560 (26.8%) died in an acute care hospital (Table 2). In total, 12 573 (51.4%) decedents were enrolled in hospice prior to death, with a median (IQR) time from sepsis discharge to hospice enrollment of 100 (36-199) days. The median (IQR) hospice length of stay was 10 (3-33) days, with 5729 (45.6%) of those enrolled receiving hospice services for 7 or fewer days prior to death.

    Factors Independently Associated With Risk of 1-Year Mortality

    In multivariate analyses, several factors were found to be independently associated with an increased risk of 1-year mortality (Table 3). Patient-level risk factors included older age (≥85 years, OR, 1.47; 95% CI, 1.40-1.54; P < .001) and the presence of comorbid conditions in general and cancer diagnosis in particular (OR, 3.66; 95% CI, 3.50-3.83; P < .001). Sepsis-related factors known at the time of discharge that were independently associated with an increased risk of 1-year mortality included severe sepsis (OR, 1.30; 95% CI, 1.23-1.37; P < .001), pneumonia and other respiratory infection source (OR, 1.14; 95% CI, 1.09-1.18; P < .001), and ICU use (OR, 1.07; 95% CI, 1.03-1.11; P < .001). Characteristics that appeared to be protective against mortality within the year after sepsis were age younger than 65 years, female sex, obesity, hypertension, and postoperative infection, or a surgical admission during the index sepsis admission.

    The initial home health care assessment identified several additional factors that were independently associated with an increased risk of 1-year mortality, including dependence in multiple ADLs/IADLs (OR, 2.80; 95% CI, 2.57-3.05; P < .001), dyspnea at rest (OR, 1.53; 95% CI, 1.42-1.66; P < .001), 2 or more hospitalizations in the past 12 months (OR, 1.21; 95% CI, 1.17-1.26; P < .001), frailty (OR, 1.07; 95% CI, 1.03-1.11; P < .001), living in an assisted living setting (OR, 1.21; 95% CI, 1.12-1.31; P < .001), and an overall poor health status (OR, 2.21; 95% CI, 2.01-2.44; P < .001).

    Factors Independently Associated With Hospice Use

    Several sociodemographic and clinical factors were independently associated with higher odds of receiving hospice care prior to death, including older age (≥75 years: OR, 1.23; 95% CI, 1.14-1.32; ≥85 years: OR, 1.49; 95% CI, 1.37-1.61; P < .001), female sex (OR, 1.11; 95% CI, 1.05-1.17; P < .001), and a cancer diagnosis (OR, 2.25; 95% CI, 2.11-2.41; P < .001). Similar to the mortality model, the postdischarge initial home health assessment identified additional factors independently associated with increased odds of hospice use independent of patient-related and sepsis-related factors, including an overall poor health status (OR, 1.40; 95% CI, 1.19-1.65 P < .001) and living with someone (OR, 1.16; 95% CI, 1.08-1.26; P < .001) or in an assisted living setting (OR, 1.93; 95% CI, 1.69-2.19; P < .001) (Table 3). Of note, multiple factors were independently associated with significantly reduced odds of hospice use, including age younger than 65 years (OR, 0.76; 95% CI, 0.69-0.84; P < .001), non-Hispanic white race (OR, 0.77; 95% CI, 0.72-0.82), being Medicaid-eligible (OR, 0.77; 95% CI, 0.72-0.82; P < .001), noncancer comorbidities (chronic pulmonary disease, heart failure, diabetes, obesity, and peripheral vascular disease), ICU use (OR, 0.91; 95% CI, 0.86-0.96; P < .001), or a bone/joint/skin/tissue infection source (OR, 0.78; 95% CI, 0.72-0.85; P < .001) during the index sepsis stay.

    Role of Cancer in Mortality and Hospice Use After Sepsis

    In this cohort, 11 427 (13.0%) sepsis survivors discharged to home health care services had a cancer diagnosis. Among them, older age (≥85 years: OR, 1.09; 95% CI, 0.95-1.23; P = .21) and the source and severity of sepsis were no longer significantly associated with 1-year mortality. Of note, dependence in multiple ADLs/IADLs (OR, 2.55; 95% CI, 2.01-3.23; P < .001) and an overall poor health status (OR, 2.98; 95% CI, 2.33-3.80; P < .001) after discharge remained independently associated. Among the 5965 (24.4%) decedents with cancer, age and sex were no longer associated with hospice use. Complete results from stratified analyses are given in eTables 3 and 4 in the Supplement.


    In this national cohort study, we examined risk factors for long-term mortality and patterns of end-of-life care among sepsis survivors who were Medicare beneficiaries and were discharged to home health care. More than one-quarter of patients died in the year following discharge, with most deaths occurring within 6 months. Two-thirds of the decedents were admitted to a hospital in the last 30 days of life where more than half received care in an ICU and 1 in 4 died. We also identified several patient-related, sepsis-related, and home health assessment factors associated with mortality and hospice use after sepsis survivorship.

    The 1-year mortality rate among sepsis survivors in the present study is within the range previously reported in the literature.9,16,36 Shankar-Hari et al16 and Yende et al36 recently reported 1-year mortality rates of 15% and 17.6%, respectively, among relatively young and previously healthy patients with few comorbidities and high rates of prehospital functional independence. In contrast, Prescott et al9 found a 1-year mortality rate of 48.5% among sepsis survivors in the Health and Retirement Study cohort, which includes an older population with multiple comorbidities and at least some functional dependence at baseline. The most likely explanation for such a wide range in mortality rates is the variation in the populations studied. Although patterns over time in the risk of long-term mortality after sepsis have not been described, if the pattern follows the decline seen in acute sepsis mortality over the past decade,2 the more contemporary data used in this study may further explain the lower mortality rate observed compared with that observed by Prescott et al.9 Finally, the present study focused on the Medicare home health care population, which tends to be older and sicker and more likely to live below the federal poverty level compared with general Medicare beneficiaries.37 However, whether and how much home health care is associated with long-term sepsis survivorship remains unknown and warrants further study.

    The present study’s results support previous reports of older age, male sex, medical admission type, comorbidities, and cancer in particular, being important risk factors for long-term mortality after sepsis.16,38 Additional sepsis-related risk factors were identified, including severe sepsis, respiratory infection source, and ICU use. Such information may be useful to guide future efforts to develop and test risk stratification models among sepsis survivors, which may facilitate tailoring postdischarge care decisions. For example, the optimal intensity and timing of postacute care services39 and follow-up with primary care clinicians40 for sepsis survivors is unknown.

    Several factors were identified on the home health care assessment after discharge that were associated with death within 1 year, independent of the foregoing patient-related and sepsis-related factors. Such factors included dyspnea at rest, 2 or more hospitalizations in the past 12 months, living in an assisted living setting, and an overall poor health status. In addition, similar to recent findings among Medicare sepsis survivors discharged to a skilled nursing facility,41 dependence in multiple ADLs was an independent risk factor for mortality after sepsis in this home health care population. Moreover, we found a high prevalence of uncontrolled pain and dyspnea after discharge, 2 of the most common reasons for emergency department visits and readmissions among chronically ill patients.42,43 These findings suggest that there is a unique opportunity for trained home health care clinicians to identify high-risk sepsis survivors and facilitate targeted interventions. For example, such patients may benefit from more frequent contact with their primary care or specialty clinicians, advance care planning, palliative care consultation, or even hospice referral in some cases. Patients referred to hospice from nonhospital sources are more likely to receive end-of-life care consistent with the preferences of most patients in the United States facing serious illness, including continuous home hospice care and dying at home.29,44

    Among the decedents in this cohort, we found that approximately two-thirds were admitted to the ICU in the last 30 days of life, which is a rate nearly 3 times that recently reported among all Medicare fee-for-service beneficiaries during a similar time frame.45 Furthermore, ICU use during the index sepsis stay was associated with lower odds of subsequent hospice use prior to death despite being associated with significantly increased risk of mortality. Although this observational study could not determine causality, we believe that prior ICU use as a potential barrier to hospice enrollment is an important area for further exploration. It is possible that for patients who recently survived a sepsis hospitalization and often an ICU stay, patients, families, and clinicians alike may rely on that past performance to predict the future.46 Such performance heuristics are common in medical decision-making, and often serve as a barrier to seeing the overall trajectory of functional decline that is common among patients with chronic illness.47

    Although the rate of hospice enrollment among the decedents in this home health sepsis cohort was similar to a recent report among a general Medicare fee-for-service population,45 the median hospice length of stay in the present study was considerably shorter. This finding suggests a missed opportunity within home health care to improve end-of-life care, supported by the finding that the median time to hospice admission was 100 days from sepsis discharge. For example, patients with a diagnosis of heart failure, chronic pulmonary disease, or peripheral vascular disease in this cohort had considerably increased mortality risk, yet they were less likely to receive hospice care. Thus, earlier disease-specific interventions to improve end-of-life care in this population may be needed.


    This study has limitations. First, although this study offers an important first look at mortality risk and end-of-life care outcomes among sepsis survivors discharged to home health care, these results may not be generalizable to other sepsis populations, such as those discharged to home without home health care or those discharged to institutional postacute care. Prospective observational studies, designed to confirm whether functional and overall health status assessment are associated with mortality among sepsis survivors discharged to home, are needed. Second, given the inherent limitations of identifying sepsis survivors using administrative claims, future studies may benefit from sepsis identification from electronic clinical records in accordance with current international sepsis definitions.48 Third, a comparison group of nonsepsis hospital discharges to home health care was not available to quantify how many of the present study’s findings are directly attributable to sepsis vs other diagnoses with home health care. We were also unable to examine in this retrospective study whether the outcomes were mediated by types of home health care services received owing to unmeasurable indication bias. Prior studies have found mixed evidence for the efficacy of physical or occupational therapy after sepsis49,50; consequently, pragmatic randomized trials are needed. In addition, we acknowledge that patient preferences regarding end-of-life care were unknown for this cohort such that it was not possible to assess whether hospitalization near the end of life or dying in the hospital reflected goal-concordant care in some cases.51


    Improvements in sepsis care have led to an increase in short-term survival, yet long-term mortality rates after hospital discharge remain high. Many sepsis survivors have readily identifiable characteristics that are associated with an increased risk of death, which may help direct interventions that mitigate the high rates of aggressive and intensive care experienced near the end of life among this population. Further research is needed to understand the association of postacute care services with mortality risk and end-of-life outcomes among sepsis survivors.

    Back to top
    Article Information

    Accepted for Publication: December 27, 2019.

    Published: February 26, 2020. doi:10.1001/jamanetworkopen.2020.0038

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Courtright KR et al. JAMA Network Open.

    Corresponding Author: Katherine R. Courtright, MD, MS, Blockley Hall 303, 423 Guardian Dr, Philadelphia, PA 19104 (katherine.courtright@pennmedicine.upenn.edu).

    Author Contributions: Ms Jordan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Courtright, Jordan, Deb, Mikkelsen.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Courtright, Jordan, Mikkelsen.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Courtright, Jordan, Barrón, Moore, Mikkelsen.

    Obtained funding: Murtaugh, Bowles, Mikkelsen.

    Administrative, technical, or material support: Courtright, Jordan, Murtaugh, Moore, Bowles.

    Supervision: Murtaugh, Mikkelsen.

    Conflict of Interest Disclosures: Dr Murtaugh and Ms Barrón reported receiving grants from the National Institute of Nursing Research (NINR) of the National Institutes of Health (NIH) during the conduct of the study. Dr Deb reported receiving grants from the NIH during the conduct of the study. Dr Mikkelsen reported receiving grants from the NINR of the NIH during the conduct of the study and consulting fees from the Hospital and Healthsystem Association of Pennsylvania outside the submitted work. No other disclosures were reported.

    Funding/Support: Funding was provided by the National Institute of Nursing Research at the National Institutes of Health (grant No. 1R01NR016014).

    Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Iwashyna  TJ, Cooke  CR, Wunsch  H, Kahn  JM.  Population burden of long-term survivorship after severe sepsis in older Americans.  J Am Geriatr Soc. 2012;60(6):1070-1077. doi:10.1111/j.1532-5415.2012.03989.xPubMedGoogle ScholarCrossref
    Meyer  N, Harhay  MO, Small  DS,  et al.  Temporal trends in incidence, sepsis-related mortality, and hospital-based acute care after sepsis.  Crit Care Med. 2018;46(3):354-360. doi:10.1097/CCM.0000000000002872PubMedGoogle ScholarCrossref
    Gaieski  DF, Edwards  JM, Kallan  MJ, Carr  BG.  Benchmarking the incidence and mortality of severe sepsis in the United States.  Crit Care Med. 2013;41(5):1167-1174. doi:10.1097/CCM.0b013e31827c09f8PubMedGoogle ScholarCrossref
    Iwashyna  TJ, Ely  EW, Smith  DM, Langa  KM.  Long-term cognitive impairment and functional disability among survivors of severe sepsis.  JAMA. 2010;304(16):1787-1794. doi:10.1001/jama.2010.1553PubMedGoogle ScholarCrossref
    Jones  TK, Fuchs  BD, Small  DS,  et al.  Post-acute care use and hospital readmission after sepsis.  Ann Am Thorac Soc. 2015;12(6):904-913. doi:10.1513/AnnalsATS.201411-504OCPubMedGoogle ScholarCrossref
    Ortego  A, Gaieski  DF, Fuchs  BD,  et al.  Hospital-based acute care use in survivors of septic shock.  Crit Care Med. 2015;43(4):729-737. doi:10.1097/CCM.0000000000000693PubMedGoogle ScholarCrossref
    Prescott  HC, Langa  KM, Liu  V, Escobar  GJ, Iwashyna  TJ.  Increased 1-year healthcare use in survivors of severe sepsis.  Am J Respir Crit Care Med. 2014;190(1):62-69. doi:10.1164/rccm.201403-0471OCPubMedGoogle ScholarCrossref
    Dietz  BW, Jones  TK, Small  DS, Gaieski  DF, Mikkelsen  ME.  The relationship between index hospitalizations, sepsis, and death or transition to hospice during 30-day hospital readmissions.  Med Care. 2017;55(4):362-370. doi:10.1097/MLR.0000000000000669PubMedGoogle ScholarCrossref
    Prescott  HC, Sjoding  MW, Langa  KM, Iwashyna  TJ, McAuley  DF.  Late mortality after acute hypoxic respiratory failure.  Thorax. 2017;thoraxjnl-2017-210109. doi:10.1136/thoraxjnl-2017-210109PubMedGoogle Scholar
    Quartin  AA, Schein  RM, Kett  DH, Peduzzi  PN; Department of Veterans Affairs Systemic Sepsis Cooperative Studies Group.  Magnitude and duration of the effect of sepsis on survival.  JAMA. 1997;277(13):1058-1063. doi:10.1001/jama.1997.03540370048035PubMedGoogle ScholarCrossref
    Wang  HE, Szychowski  JM, Griffin  R, Safford  MM, Shapiro  NI, Howard  G.  Long-term mortality after community-acquired sepsis: a longitudinal population-based cohort study.  BMJ Open. 2014;4(1):e004283. doi:10.1136/bmjopen-2013-004283PubMedGoogle Scholar
    Maley  JH, Mikkelsen  ME.  Short-term gains with long-term consequences: the evolving story of sepsis survivorship.  Clin Chest Med. 2016;37(2):367-380. doi:10.1016/j.ccm.2016.01.017PubMedGoogle ScholarCrossref
    Prescott  HC, Angus  DC.  Enhancing recovery from sepsis: a review.  JAMA. 2018;319(1):62-75. doi:10.1001/jama.2017.17687PubMedGoogle ScholarCrossref
    HCUP website. An all-payer view of hospital discharge to postacute care, 2013. https://hcup-us.ahrq.gov/reports/statbriefs/sb205-Hospital-Discharge-Postacute-Care.jsp. Accessed December 19, 2019.
    Lee  JT, Mikkelsen  ME, Qi  M, Werner  RM.  Trends in post-acute care use after admissions for sepsis.  Ann Am Thorac Soc.2020;17(1):118-121. doi:10.1513/AnnalsATS.201905-368RLPubMedGoogle Scholar
    Shankar-Hari  M, Harrison  DA, Ferrando-Vivas  P, Rubenfeld  GD, Rowan  K.  Risk factors at index hospitalization associated with longer-term mortality in adult sepsis survivors.  JAMA Netw Open. 2019;2(5):e194900. doi:10.1001/jamanetworkopen.2019.4900PubMedGoogle Scholar
    Institute of Medicine.  Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: National Academies Press; 2014.
    Teno  JM, Gozalo  PL, Bynum  JP,  et al.  Change in end-of-life care for Medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009.  JAMA. 2013;309(5):470-477. doi:10.1001/jama.2012.207624PubMedGoogle ScholarCrossref
    Naylor  MD, Aiken  LH, Kurtzman  ET, Olds  DM, Hirschman  KB.  The care span: the importance of transitional care in achieving health reform.  Health Aff (Millwood). 2011;30(4):746-754. doi:10.1377/hlthaff.2011.0041PubMedGoogle ScholarCrossref
    Meier  DE, Back  AL, Berman  A, Block  SD, Corrigan  JM, Morrison  RS.  A national strategy for palliative care.  Health Aff (Millwood). 2017;36(7):1265-1273. doi:10.1377/hlthaff.2017.0164PubMedGoogle ScholarCrossref
    Ferrell  BR, Twaddle  ML, Melnick  A, Meier  DE.  National Consensus Project Clinical Practice Guidelines for Quality Palliative Care Guidelines, 4th Edition.  J Palliat Med. 2018;21(12):1684-1689. doi:10.1089/jpm.2018.0431Google Scholar
    Rietjens  JAC, Sudore  RL, Connolly  M,  et al; European Association for Palliative Care.  Definition and recommendations for advance care planning: an international consensus supported by the European Association for Palliative Care.  Lancet Oncol. 2017;18(9):e543-e551. doi:10.1016/S1470-2045(17)30582-XPubMedGoogle ScholarCrossref
    Kinatukara  S, Rosati  RJ, Huang  L.  Assessment of OASIS reliability and validity using several methodological approaches.  Home Health Care Serv Q. 2005;24(3):23-38. doi:10.1300/J027v24n03_02PubMedGoogle ScholarCrossref
    O’Connor  M, Davitt  JK.  The Outcome and Assessment Information Set (OASIS): a review of validity and reliability.  Home Health Care Serv Q. 2012;31(4):267-301. doi:10.1080/01621424.2012.703908PubMedGoogle ScholarCrossref
    von Elm  E, Altman  DG, Egger  M, Pocock  SJ, Gøtzsche  PC, Vandenbroucke  JP; STROBE Initiative.  The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.  Ann Intern Med. 2007;147(8):573-577. doi:10.7326/0003-4819-147-8-200710160-00010PubMedGoogle ScholarCrossref
    Whittaker  SA, Mikkelsen  ME, Gaieski  DF, Koshy  S, Kean  C, Fuchs  BD.  Severe sepsis cohorts derived from claims-based strategies appear to be biased toward a more severely ill patient population.  Crit Care Med. 2013;41(4):945-953. doi:10.1097/CCM.0b013e31827466f1PubMedGoogle ScholarCrossref
    Angus  DC, Linde-Zwirble  WT, Lidicker  J, Clermont  G, Carcillo  J, Pinsky  MR.  Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care.  Crit Care Med. 2001;29(7):1303-1310. doi:10.1097/00003246-200107000-00002PubMedGoogle ScholarCrossref
    Iwashyna  TJ, Odden  A, Rohde  J,  et al.  Identifying patients with severe sepsis using administrative claims: patient-level validation of the Angus implementation of the international consensus conference definition of severe sepsis.  Med Care. 2014;52(6):e39-e43. doi:10.1097/MLR.0b013e318268ac86PubMedGoogle ScholarCrossref
    Miller  SC, Kinzbrunner  B, Pettit  P, Williams  JR.  How does the timing of hospice referral influence hospice care in the last days of life?  J Am Geriatr Soc. 2003;51(6):798-806. doi:10.1046/j.1365-2389.2003.51253.xPubMedGoogle ScholarCrossref
    Rickerson  E, Harrold  J, Kapo  J, Carroll  JT, Casarett  D.  Timing of hospice referral and families’ perceptions of services: are earlier hospice referrals better?  J Am Geriatr Soc. 2005;53(5):819-823. doi:10.1111/j.1532-5415.2005.53259.xPubMedGoogle ScholarCrossref
    Schockett  ER, Teno  JM, Miller  SC, Stuart  B.  Late referral to hospice and bereaved family member perception of quality of end-of-life care.  J Pain Symptom Manage. 2005;30(5):400-407. doi:10.1016/j.jpainsymman.2005.04.013PubMedGoogle ScholarCrossref
    Dumanovsky  T, Augustin  R, Rogers  M, Lettang  K, Meier  DE, Morrison  RS.  The growth of palliative care in U.S. hospitals: a status report.  J Palliat Med. 2016;19(1):8-15. doi:10.1089/jpm.2015.0351PubMedGoogle ScholarCrossref
    Murtaugh  CM, Deb  P, Zhu  C,  et al.  Reducing readmissions among heart failure patients discharged to home health care: effectiveness of early and intensive nursing services and early physician follow-up.  Health Serv Res. 2017;52(4):1445-1472. doi:10.1111/1475-6773.12537PubMedGoogle ScholarCrossref
    Belsley  DA, Kuh  E, Welsch  RE.  Regression Diagnostics: Identifying Influential Data and Sources of Collinearity. New York, NY: Wiley; 1980. doi:10.1002/0471725153
    Elixhauser  A, Steiner  C, Harris  DR, Coffey  RM.  Comorbidity measures for use with administrative data.  Med Care. 1998;36(1):8-27. doi:10.1097/00005650-199801000-00004PubMedGoogle ScholarCrossref
    Yende  S, Kellum  JA, Talisa  VB,  et al.  Long-term host immune response trajectories among hospitalized patients with sepsis.  JAMA Netw Open. 2019;2(8):e198686. doi:10.1001/jamanetworkopen.2019.8686PubMedGoogle Scholar
    Avalere Health. Home Health Chartbook 2017: Prepared for the Alliance for Home Health Quality and Innovation. https://www.ahhqi.org/research/home-health-chartbook. Published March 2017. Accessed January 22, 2020.
    Francisco  J, Aragão  I, Cardoso  T.  Risk factors for long-term mortality in patients admitted with severe infection.  BMC Infect Dis. 2018;18(1):161. doi:10.1186/s12879-018-3054-4PubMedGoogle ScholarCrossref
    Bowles  KH, Ratcliffe  SJ, Holmes  JH,  et al.  Using a decision support algorithm for referrals to post-acute care.  J Am Med Dir Assoc. 2019;20(4):408-413. doi:10.1016/j.jamda.2018.08.016PubMedGoogle ScholarCrossref
    Deb  P, Murtaugh  CM, Bowles  KH,  et al.  Does early follow-up improve the outcomes of sepsis survivors discharged to home health care?  Med Care. 2019;57(8):633-640. doi:10.1097/MLR.0000000000001152PubMedGoogle ScholarCrossref
    Ehlenbach  WJ, Gilmore-Bykovskyi  A, Repplinger  MD,  et al.  sepsis survivors admitted to skilled nursing facilities: cognitive impairment, activities of daily living dependence, and survival.  Crit Care Med. 2018;46(1):37-44. doi:10.1097/CCM.0000000000002755PubMedGoogle ScholarCrossref
    Mayer  DK, Travers  D, Wyss  A, Leak  A, Waller  A.  Why do patients with cancer visit emergency departments? results of a 2008 population study in North Carolina.  J Clin Oncol. 2011;29(19):2683-2688. doi:10.1200/JCO.2010.34.2816PubMedGoogle ScholarCrossref
    Raghavan  D, Bartter  T, Joshi  M.  How to reduce hospital readmissions in chronic obstructive pulmonary disease?  Curr Opin Pulm Med. 2016;22(2):106-112. doi:10.1097/MCP.0000000000000245PubMedGoogle ScholarCrossref
    Higginson  IJ, Sen-Gupta  GJ.  Place of care in advanced cancer: a qualitative systematic literature review of patient preferences.  J Palliat Med. 2000;3(3):287-300. doi:10.1089/jpm.2000.3.287PubMedGoogle ScholarCrossref
    Teno  JM, Gozalo  P, Trivedi  AN,  et al.  Site of death, place of care, and health care transitions among US Medicare beneficiaries, 2000-2015.  JAMA. 2018;320(3):264-271. doi:10.1001/jama.2018.8981PubMedGoogle ScholarCrossref
    Critcher  CR, Rosenzweig  EL.  The performance heuristic: a misguided reliance on past success when predicting prospects for improvement.  J Exp Psychol Gen. 2014;143(2):480-485. doi:10.1037/a0034129PubMedGoogle ScholarCrossref
    Lunney  JR, Lynn  J, Foley  DJ, Lipson  S, Guralnik  JM.  Patterns of functional decline at the end of life.  JAMA. 2003;289(18):2387-2392. doi:10.1001/jama.289.18.2387PubMedGoogle ScholarCrossref
    Rhee  C, Kadri  S, Huang  SS,  et al.  Objective sepsis surveillance using electronic clinical data.  Infect Control Hosp Epidemiol. 2016;37(2):163-171. doi:10.1017/ice.2015.264PubMedGoogle ScholarCrossref
    Chao  PW, Shih  CJ, Lee  YJ,  et al.  Association of postdischarge rehabilitation with mortality in intensive care unit survivors of sepsis.  Am J Respir Crit Care Med. 2014;190(9):1003-1011. doi:10.1164/rccm.201406-1170OCPubMedGoogle ScholarCrossref
    Elliott  D, McKinley  S, Alison  J,  et al.  Health-related quality of life and physical recovery after a critical illness: a multi-centre randomised controlled trial of a home-based physical rehabilitation program.  Crit Care. 2011;15(3):R142. doi:10.1186/cc10265PubMedGoogle ScholarCrossref
    Halpern  SD.  Goal-concordant care—searching for the Holy Grail.  N Engl J Med. 2019;381(17):1603-1606. doi:10.1056/NEJMp1908153PubMedGoogle ScholarCrossref
    Centers for Medicare & Medicaid. Research Data Assistance Center. https://www.resdac.org/. Accessed January 15, 2020.