[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.
Flow Diagram for Identification of Sepsis Survivor Cohort
Flow Diagram for Identification of Sepsis Survivor Cohort

Sepsis survivors were defined as adult patients who survived to hospital discharge following an index critical care admission with sepsis. We defined the index critical care admission for sepsis as the hospitalization in the study period that included the first critical care unit admission for sepsis as defined by the Third International Consensus Definitions for Sepsis and Septic Shock.1,3 Thus, each sepsis survivor was included only once in the study, based on his or her index critical care admission for sepsis. Some patients have 1 or more critical care admissions with and without sepsis during the study period. Patients with nonsepsis admission may be readmitted during the study period for their index (first) critical care admission for sepsis.

Figure 2.
Cumulative All-Cause Long-term Mortality in Adult Sepsis Survivors Compared With Age- and Sex-Matched General Population
Cumulative All-Cause Long-term Mortality in Adult Sepsis Survivors Compared With Age- and Sex-Matched General Population

Cumulative all-cause long-term mortality in the entire cohort of sepsis survivors (as defined by the Third International Consensus Definitions for Sepsis and Septic Shock) compared with age-, sex-, and index sepsis admission year–matched general population expected probabilities of death in England, using relative survival framework (eTable 3 in the Supplement).

Figure 3.
Cumulative All-Cause Long-term Mortality by Risk Factors in Adult Sepsis Survivors
Cumulative All-Cause Long-term Mortality by Risk Factors in Adult Sepsis Survivors

A-F, Stratum-level differences in crude mortality of sepsis survivors (as defined by the Third International Consensus Definitions for Sepsis and Septic Shock) over follow-up within generic characteristics (age, ≥1 severe comorbidities, and Acute Physiology and Chronic Health Evaluation II (APACHE II) acute physiology score quartile strata) and sepsis-specific characteristics (site of infection; number of organ dysfunction; and septic shock status) (all P < .001 using Peto log-rank test). Number of patients at risk is shown in eTable 4 in the Supplement.

Figure 4.
Adjusted Hazard Ratios (HRs) for Each Risk Factor on Long-term Mortality by Survival Time in Adult Sepsis Survivors
Adjusted Hazard Ratios (HRs) for Each Risk Factor on Long-term Mortality by Survival Time in Adult Sepsis Survivors

Adjusted HRs generated using the primary Cox regression model (eTable 2 in the Supplement) are shown for generic risk factors, time-varying coefficients (for age in decades, acute physiology component of Acute Physiology and Chronic Health Evaluation II [APACHE II] score per 5-point increments, and severe comorbidity) and interactions (A) and sepsis-specific risk factors (B). The P values represent test for homogeneity within groups, estimated using a postestimation command to look for significant differences after the Cox regression model. See the Methods section for the base category for each risk factor. Sepsis was defined according to the criteria of the Third International Consensus Definitions for Sepsis and Septic Shock. APII-APS indicates acute physiology component of APACHE II; and LOS, length of stay.

Table.  
Characteristics of Sepsis Survivor Cohorta
Characteristics of Sepsis Survivor Cohorta
1.
Singer  M, Deutschman  CS, Seymour  CW,  et al.  The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).  JAMA. 2016;315(8):801-810. doi:10.1001/jama.2016.0287PubMedGoogle ScholarCrossref
2.
Fleischmann  C, Scherag  A, Adhikari  NK,  et al; International Forum of Acute Care Trialists.  Assessment of global incidence and mortality of hospital-treated sepsis: current estimates and limitations.  Am J Respir Crit Care Med. 2016;193(3):259-272. doi:10.1164/rccm.201504-0781OCPubMedGoogle ScholarCrossref
3.
Shankar-Hari  M, Harrison  DA, Rubenfeld  GD, Rowan  K.  Epidemiology of sepsis and septic shock in critical care units: comparison between sepsis-2 and sepsis-3 populations using a national critical care database.  Br J Anaesth. 2017;119(4):626-636. doi:10.1093/bja/aex234PubMedGoogle ScholarCrossref
4.
Kaukonen  KM, Bailey  M, Suzuki  S, Pilcher  D, Bellomo  R.  Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012.  JAMA. 2014;311(13):1308-1316. doi:10.1001/jama.2014.2637PubMedGoogle ScholarCrossref
5.
Prescott  HC, Angus  DC.  Enhancing recovery from sepsis: a review.  JAMA. 2018;319(1):62-75. doi:10.1001/jama.2017.17687PubMedGoogle ScholarCrossref
6.
Shankar-Hari  M, Ambler  M, Mahalingasivam  V, Jones  A, Rowan  K, Rubenfeld  GD.  Evidence for a causal link between sepsis and long-term mortality: a systematic review of epidemiologic studies.  Crit Care. 2016;20(1):101. doi:10.1186/s13054-016-1276-7PubMedGoogle ScholarCrossref
7.
Ou  SM, Chu  H, Chao  PW,  et al.  Long-term mortality and major adverse cardiovascular events in sepsis survivors. a nationwide population-based study.  Am J Respir Crit Care Med. 2016;194(2):209-217. doi:10.1164/rccm.201510-2023OCPubMedGoogle ScholarCrossref
8.
Prescott  HC, Osterholzer  JJ, Langa  KM, Angus  DC, Iwashyna  TJ.  Late mortality after sepsis: propensity matched cohort study.  BMJ. 2016;353:i2375. doi:10.1136/bmj.i2375PubMedGoogle ScholarCrossref
9.
Davis  JS, He  V, Anstey  NM, Condon  JR.  Long term outcomes following hospital admission for sepsis using relative survival analysis: a prospective cohort study of 1,092 patients with 5 year follow up.  PLoS One. 2014;9(12):e112224. doi:10.1371/journal.pone.0112224PubMedGoogle ScholarCrossref
10.
Shankar-Hari  M, Rubenfeld  GD.  Understanding long-term outcomes following sepsis: implications and challenges.  Curr Infect Dis Rep. 2016;18(11):37. doi:10.1007/s11908-016-0544-7PubMedGoogle ScholarCrossref
11.
Shankar-Hari  M, Harrison  DA, Rowan  KM.  Differences in impact of definitional elements on mortality precludes international comparisons of sepsis epidemiology-a cohort study illustrating the need for standardized reporting.  Crit Care Med. 2016;44(12):2223-2230. doi:10.1097/CCM.0000000000001876PubMedGoogle ScholarCrossref
12.
Vandenbroucke  JP, von Elm  E, Altman  DG,  et al; STROBE Initiative.  Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration.  PLoS Med. 2007;4(10):e297. doi:10.1371/journal.pmed.0040297PubMedGoogle ScholarCrossref
13.
Young  JD, Goldfrad  C, Rowan  K.  Development and testing of a hierarchical method to code the reason for admission to intensive care units: the ICNARC Coding Method.  Br J Anaesth. 2001;87(4):543-548. doi:10.1093/bja/87.4.543PubMedGoogle ScholarCrossref
14.
Harrison  DA, Brady  AR, Rowan  K.  Case mix, outcome and length of stay for admissions to adult, general critical care units in England, Wales and Northern Ireland: the Intensive Care National Audit & Research Centre Case Mix Programme Database.  Crit Care. 2004;8(2):R99-R111. doi:10.1186/cc2834PubMedGoogle ScholarCrossref
15.
NHS Digital. Linked mortality data from Hospital Episode Statistics and the Office for National Statistics. http://content.digital.nhs.uk/hesonsmortality. Accessed April 4, 2017.
16.
Shankar-Hari  M, Phillips  GS, Levy  ML,  et al; Sepsis Definitions Task Force.  Developing a new definition and assessing new clinical criteria for septic shock: for the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).  JAMA. 2016;315(8):775-787. doi:10.1001/jama.2016.0289PubMedGoogle ScholarCrossref
17.
Knaus  WA, Draper  EA, Wagner  DP, Zimmerman  JE.  APACHE II: a severity of disease classification system.  Crit Care Med. 1985;13(10):818-829. doi:10.1097/00003246-198510000-00009PubMedGoogle ScholarCrossref
18.
Grambsch  PM, Therneau  TM.  Proportional hazards tests and diagnostics based on weighted residuals.  Biometrika. 1994;81(3):515-526. doi:10.1093/biomet/81.3.515Google ScholarCrossref
19.
Austin  PC.  A tutorial on multilevel survival analysis: methods, models and applications.  Int Stat Rev. 2017;85(2):185-203. doi:10.1111/insr.12214PubMedGoogle ScholarCrossref
20.
Dickman  PW, Coviello  E.  Estimating and modeling relative survival.  Stata J. 2015;15(1):186-215. doi:10.1177/1536867X1501500112Google ScholarCrossref
21.
Office for National Statistics. https://www.ons.gov.uk. Accessed February 28, 2017.
22.
Scicluna  BP, van Vught  LA, Zwinderman  AH,  et al; MARS consortium.  Classification of patients with sepsis according to blood genomic endotype: a prospective cohort study.  Lancet Respir Med. 2017;5(10):816-826. doi:10.1016/S2213-2600(17)30294-1PubMedGoogle ScholarCrossref
23.
Burnham  KL, Davenport  EE, Radhakrishnan  J,  et al.  Shared and distinct aspects of the sepsis transcriptomic response to fecal peritonitis and pneumonia.  Am J Respir Crit Care Med. 2017;196(3):328-339. doi:10.1164/rccm.201608-1685OCPubMedGoogle ScholarCrossref
24.
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
25.
Yende  S, D’Angelo  G, Kellum  JA,  et al; GenIMS Investigators.  Inflammatory markers at hospital discharge predict subsequent mortality after pneumonia and sepsis.  Am J Respir Crit Care Med. 2008;177(11):1242-1247. doi:10.1164/rccm.200712-1777OCPubMedGoogle ScholarCrossref
26.
Prescott  HC, Langa  KM, Iwashyna  TJ.  Readmission diagnoses after hospitalization for severe sepsis and other acute medical conditions.  JAMA. 2015;313(10):1055-1057. doi:10.1001/jama.2015.1410PubMedGoogle ScholarCrossref
27.
Corrales-Medina  VF, Alvarez  KN, Weissfeld  LA,  et al.  Association between hospitalization for pneumonia and subsequent risk of cardiovascular disease.  JAMA. 2015;313(3):264-274. doi:10.1001/jama.2014.18229PubMedGoogle ScholarCrossref
28.
Wunsch  H, Angus  DC, Harrison  DA, Linde-Zwirble  WT, Rowan  KM.  Comparison of medical admissions to intensive care units in the United States and United Kingdom.  Am J Respir Crit Care Med. 2011;183(12):1666-1673. doi:10.1164/rccm.201012-1961OCPubMedGoogle ScholarCrossref
29.
Rhodes  A, Ferdinande  P, Flaatten  H, Guidet  B, Metnitz  PG, Moreno  RP.  The variability of critical care bed numbers in Europe.  Intensive Care Med. 2012;38(10):1647-1653. doi:10.1007/s00134-012-2627-8PubMedGoogle ScholarCrossref
30.
Kellum  JA, Kong  L, Fink  MP,  et al; GenIMS Investigators.  Understanding the inflammatory cytokine response in pneumonia and sepsis: results of the Genetic and Inflammatory Markers of Sepsis (GenIMS) study.  Arch Intern Med. 2007;167(15):1655-1663. doi:10.1001/archinte.167.15.1655PubMedGoogle ScholarCrossref
31.
van der Poll  T, van de Veerdonk  FL, Scicluna  BP, Netea  MG.  The immunopathology of sepsis and potential therapeutic targets.  Nat Rev Immunol. 2017;17(7):407-420. doi:10.1038/nri.2017.36PubMedGoogle ScholarCrossref
32.
Rhodes  A, Evans  LE, Alhazzani  W,  et al.  Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016.  Intensive Care Med. 2017;43(3):304-377. doi:10.1007/s00134-017-4683-6PubMedGoogle ScholarCrossref
33.
Seymour  CW, Gesten  F, Prescott  HC,  et al.  Time to treatment and mortality during mandated emergency care for sepsis.  N Engl J Med. 2017;376(23):2235-2244. doi:10.1056/NEJMoa1703058PubMedGoogle ScholarCrossref
34.
Liu  VX, Fielding-Singh  V, Greene  JD,  et al.  The timing of early antibiotics and hospital mortality in sepsis.  Am J Respir Crit Care Med. 2017;196(7):856-863. doi:10.1164/rccm.201609-1848OCPubMedGoogle ScholarCrossref
35.
Riché  F, Chousterman  BG, Valleur  P, Mebazaa  A, Launay  JM, Gayat  E.  Protracted immune disorders at one year after ICU discharge in patients with septic shock.  Crit Care. 2018;22(1):42. doi:10.1186/s13054-017-1934-4PubMedGoogle ScholarCrossref
36.
Arens  C, Bajwa  SA, Koch  C,  et al.  Sepsis-induced long-term immune paralysis—results of a descriptive, explorative study.  Crit Care. 2016;20(1):93. doi:10.1186/s13054-016-1233-5PubMedGoogle ScholarCrossref
37.
DeMerle  KM, Royer  SC, Mikkelsen  ME, Prescott  HC.  Readmissions for recurrent sepsis: new or relapsed infection?  Crit Care Med. 2017;45(10):1702-1708. doi:10.1097/CCM.0000000000002626PubMedGoogle ScholarCrossref
38.
Shen  E, Koyama  SY, Huynh  DN,  et al.  Association of a dedicated post-hospital discharge follow-up visit and 30-day readmission risk in a Medicare Advantage population.  JAMA Intern Med. 2017;177(1):132-135. doi:10.1001/jamainternmed.2016.7061PubMedGoogle ScholarCrossref
40.
Shankar-Hari  M, Fear  D, Lavender  P,  et al.  Activation-associated accelerated apoptosis of memory B cells in critically ill patients with sepsis.  Crit Care Med. 2017;45(5):875-882. doi:10.1097/CCM.0000000000002380PubMedGoogle ScholarCrossref
41.
Reinhart  K, Daniels  R, Kissoon  N, Machado  FR, Schachter  RD, Finfer  S.  Recognizing sepsis as a global health priority—a WHO resolution.  N Engl J Med. 2017;377(5):414-417. doi:10.1056/NEJMp1707170PubMedGoogle ScholarCrossref
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
    Views 4,412
    Original Investigation
    Critical Care Medicine
    May 31, 2019

    Risk Factors at Index Hospitalization Associated With Longer-term Mortality in Adult Sepsis Survivors

    Author Affiliations
    • 1Intensive Care Unit Support Offices, St Thomas' Hospital, Guy's and St Thomas' NHS Foundation Trust, London, United Kingdom
    • 2School of Immunology & Microbial Sciences, King’s College London, London, United Kingdom
    • 3Intensive Care National Audit & Research Centre, London, United Kingdom
    • 4Interdepartmental Division of Critical Care Medicine, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
    JAMA Netw Open. 2019;2(5):e194900. doi:10.1001/jamanetworkopen.2019.4900
    Key Points español 中文 (chinese)

    Question  Which generic and sepsis-specific patient characteristics, known during index critical care admission for sepsis, are independently associated with long-term mortality in sepsis survivors?

    Findings  In this cohort study of 94 748 adult sepsis survivors, age, male sex, 1 or more severe comorbidities, prehospitalization dependency, nonsurgical status, acute severity of illness, site of infection, and organ dysfunction were independently associated with long-term mortality.

    Meaning  Generic and sepsis-specific risk factors, known during index critical care admission for sepsis, could be used to identify a higher-risk sepsis survivor population for targeted strategies aimed at reducing the excess risk of long-term mortality.

    Abstract

    Importance  Sepsis survivors, defined as adult patients who survived to hospital discharge following a critical care unit admission for sepsis, are at increased risk of long-term mortality. Identifying factors independently associated with long-term mortality, known during critical care admission for sepsis, could inform targeted strategies to reduce this risk.

    Objective  To assess, in adult sepsis survivors, factors independently associated with long-term mortality, known during their index critical care admission for sepsis, meeting Third International Consensus Definitions for Sepsis and Septic Shock criteria.

    Design, Setting, and Participants  This cohort study included a nationally representative sample of 94 748 adult sepsis survivors from 192 critical care units in England. Participants were identified from consecutive critical care admissions between April 1, 2009, and March 31, 2014, with survival status ascertained as of March 31, 2015. Statistical analyses were completed in June 2017.

    Exposures  Generic patient characteristics (age, sex, ethnicity, severe comorbidities [defined using the Acute Physiology and Chronic Health Evaluation II method], dependency, surgical status, and acute illness severity [scored using the Acute Physiology and Chronic Health Evaluation II acute physiology component]) and sepsis-specific patient characteristics (site of infection, number of organ dysfunctions, and septic shock status) known during index critical care admission for sepsis.

    Main Outcomes and Measures  Long-term mortality in adult sepsis survivors with maximum follow-up of 6 years. Adjusted hazard ratios (HRs) were estimated using Cox regression for both generic and sepsis-specific patient characteristics.

    Results  Sepsis survivors had a mean (SD) age of 61.3 (17.0) years, 43 584 (46.0%) were female, and 86 056 (90.8%) were white. A total of 46.3% had respiratory site of infection. By 1 year from hospital discharge, 15% of sepsis survivors had died, with 6% to 8% dying per year over the subsequent 5 years. Age, sex, race/ethnicity, severe comorbidities, dependency, nonsurgical status, and site of infection were independently associated with long-term mortality. Compared with single-organ dysfunction, having 2 or 3 organ dysfunctions was associated with increased risk of long-term mortality (adjusted HR, 1.07; 95% CI, 1.01-1.13; and adjusted HR, 1.18; 95% CI, 1.03-1.14, respectively), while having 4 organ dysfunctions or more was not associated with increased risk. Unexpectedly, the Acute Physiology and Chronic Health Evaluation acute physiology component score had an incremental association with long-term mortality (adjusted HR, 1.11 for every 5-point increase; 95% CI, 1.08-1.13). The adjusted HR for septic shock was 0.89 (95% CI, 0.85-0.92).

    Conclusions and Relevance  This study suggests that generic and sepsis-specific risk factors, known during index critical care admission for sepsis, could identify a high-risk sepsis survivor population for biological characterization and designing interventions to reduce long-term mortality.

    Introduction

    Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection.1 Sepsis is common, with an estimated population incidence of 148 cases per 100 000 person-years.2 Sepsis epidemiology studies from the past 2 decades consistently highlight both increasing incidence and improving acute mortality.3,4 The result of this combination is an increasing number of sepsis survivors (defined in this article as adult patients who survived to hospital discharge following a critical care unit admission for sepsis). The ongoing health care needs of adult sepsis survivors are an emerging global health care challenge.5

    It is also well recognized that sepsis survivors have a greater risk of long-term mortality when compared with patients hospitalized for nonsepsis reasons and the general population.6-9 A systematic review6 that included 59 studies and explored the association between sepsis and long-term mortality indicated that, on average, 16% of sepsis survivors die in the first year following their index hospitalization for sepsis. There are major variations in the sepsis case definition and in the risk factors assessed for this long-term mortality between studies. When studying risk factors for long-term mortality in sepsis survivors, the factors associated with short-term mortality may overwhelm and disguise factors associated with long-term mortality—as observed in many sepsis epidemiology studies included in the recent systematic review.6,10 The reason for this is that an increase in sepsis severity may worsen cumulative long-term mortality by increasing hospital mortality. This competing risk of hospital mortality on long-term mortality may hide the impact of acute illness characteristics, unless we study a sepsis survivor cohort. Therefore, the following should be considered in the selection of factors to study: (1) potential factors associated with mortality should include both the generic and sepsis-specific factors that resulted in the critical care admission for sepsis11; (2) generic factors should relate to the baseline risk of death irrespective of sepsis (eg, severe comorbidity, preadmission dependency status); and (3) sepsis-specific factors should reflect the severity of sepsis. In this context, inferences are not biased when studying a sepsis survivor population (survivorship bias), as the exposure of interest is conditional on surviving to hospital discharge following a critical care admission for sepsis. If long-term mortality risk after sepsis is associated with acute illness characteristics that are known when a sepsis survivor leaves the hospital, those risk factors could be used to identify a target sepsis survivor population for follow-up care, biological characterization, and designing interventions.5,10

    In this context, we assessed which of the generic (eg, age, sex) and sepsis-specific (eg, site of infection, number of organ dysfunctions) patient characteristics, known during index critical care admission for sepsis, were independently associated with long-term mortality in a sepsis survivor population. To address our research aim, we identified a nationally representative cohort of sepsis survivors from an inception cohort of adults with a critical care admission for sepsis according to the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)3 criteria who survived to hospital discharge in England between April 1, 2009, and March 31, 2014.

    Methods
    Study Database and Approvals

    We report an observational cohort study, as per Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.12 We used data from the Intensive Care National Audit & Research Centre (ICNARC) Case Mix Programme, which is the national clinical audit covering all adult general critical care units in England. For consecutive critical care admissions, trained data collectors collect sociodemographic, comorbidity, and physiological data to precise rules and definitions for the patient’s first 24 hours following admission to critical care. Diagnostic data are determined clinically and coded using a 5-tier, hierarchical ICNARC coding method (eAppendix in the Supplement).13 Collected data undergo extensive local and central validation prior to pooling into the Case Mix Programme Database.14 Patient-level unique identifiers are also generated for additional data linkage.13 Long-term survival status and date of death were ascertained by linking the Case Mix Programme database to the Hospital Episode Statistics database and Office for National Statistics records.15 Support for the collection and use of these data has been obtained under Section 251 of the National Health Service Act 2006. Waiver of informed consent for use of this data was obtained from Wales Research Ethics Committee 5 and Confidentiality Advisory Group, which approved this study as part of the Long-term Outcomes After Critical Illness Project.

    Identification of Sepsis Survivor Cohort

    As of March 31, 2015, we ascertained the date of death for consecutive admissions to 192 adult, general critical care units in England participating in the ICNARC Case Mix Programme between April 1, 2009, and March 31, 2014. From these, we identified a cohort of Sepsis-3 sepsis survivors, defined as adult patients who survived to hospital discharge following an index critical care admission for sepsis meeting Sepsis-3 criteria. We defined the index critical care admission for sepsis as the hospitalization that included the first critical care unit admission with infection and total Sequential Organ Failure Assessment score of 2 or greater. We defined septic shock as infection, with the cardiovascular component of the Sequential Organ Failure Assessment of 2 or greater and a serum lactate concentration greater than 18 mg/dL (to convert to millimoles per liter, multiply by 0.111), as per Sepsis-3 criteria.1,16 We identified presence of infection from the reported primary (mandated) and secondary (optional) reasons for critical care admission.3 As baseline organ dysfunction at index critical care admission was unknown for our study cohort, we assumed a baseline organ dysfunction of 0, as recommended by the Sepsis-3 consensus definitions.1 We have published this operationalization,3 which included sensitivity analyses to confirm results by excluding patients with any preexisting severe comorbidity (eTable 1 in the Supplement).

    Statistical Analysis

    The primary outcome was time to death from any cause during follow-up. Generic and sepsis-specific patient characteristics, known during index critical care admission for sepsis, were considered as risk factors, informed by our systematic reviews.6,10 Generic patient characteristics included were sepsis discharge year, age, sex, ethnicity, severe comorbidities (defined using the Acute Physiology and Chronic Health Evaluation [APACHE II] method17), prehospitalization dependency, surgical status, and acute severity of illness defined using the APACHE II acute physiology component score.17 The prehospitalization dependency is a subjective assessment of whether the participant receives no, some, or total assistance to complete daily activities such as bathing, dressing, going to the toilet, moving in and out of a bed or chair, continence, and eating prior to critical care admission. Sepsis-specific patient characteristics included were site of infection, number of organ dysfunctions defined as Sequential Organ Failure Assessment score of 1 or greater per organ, and septic shock status.

    First, we generated cumulative time-to-mortality plots for all generic and sepsis-specific patient characteristics and tested equality of mortality functions using a Peto-Peto-Prentice log-rank test. The Peto-Peto-Prentice log-rank test was specifically chosen because it is not affected by differences in censoring patterns across different strata. We then assessed which generic and sepsis-specific patient characteristics were independently associated with long-term mortality in adult sepsis survivors using adjusted hazard ratios (HRs) from proportional hazards (Cox) regression modeling. The categorical variables were sex, race/ethnicity, severe comorbidity, surgical status, preadmission dependency, site of infection, number of organ dysfunctions, and septic shock status. Continuous variables were entry date (index date of discharge from hospital), age (in decades), APACHE II acute physiology component score (in 5-point increments), and index hospital length of stay. A priori, we included an interaction term between the APACHE II acute physiology component score (in 5-point increments) and severe comorbidities. We also estimated time-varying coefficients for age (in decades), APACHE II acute physiology component score (in 5-point increments), and severe comorbidities using an interaction between the covariate and follow-up time that allowed the strength of association between a risk factor and the event of interest to either increase or decrease during the course of follow-up. For all Cox regression models, we tested for evidence against the proportional hazard assumption on the basis of Schoenfeld residuals after fitting a model with Cox regression.18 We gave equal weights to different comorbidities for interpretability, accounted for clustering by intensive care units using robust standard errors, did not perform any variable selection approaches, and used the Breslow method for handling ties, which considers each event at a given time as distinct and allows all failed subjects to contribute fully to the risk set (here, failed subjects is a technical term referring to how the Breslow method handles ties within the Cox model).

    Sensitivity Analyses

    We performed 5 separate sensitivity analyses. First, to account for overlap in variables that contribute toward the APACHE II acute physiology component score and severe comorbidity (eg, creatinine, oxygenation), we performed a sensitivity analysis replicating the Cox regression model in patients without severe comorbidity. Second, to ascertain whether changes observed were influenced by coverage (additional critical care units participating in the Case Mix Programme over time), we repeated all the analyses on a subpopulation of 117 critical care units that contributed data to the Case Mix Programme for the entire study period. Third, to ascertain whether the primary analyses hold true, as in a 5-year cohort study the adjusted HR could change, we report the adjusted HR overlapping time intervals 0 to 1, 0 to 2, 0 to 3, 0 to 4, and 0 to 5 years following hospital discharge. Fourth, to ascertain whether changes observed were influenced by an unobserved random factor between critical care units that modifies the hazard function of an individual or a group or cluster of individuals, we report a mixed-effects Cox model (shared frailty model).19 Fifth, to provide direct comparison with recently published cohort studies,7-9 we estimated the excess long-term mortality in sepsis survivors compared with age-, sex-, and index year–matched expected probabilities of death20 in the general population of England between 2009 and 201421 using the relative survival framework with Ederer II estimation.

    Reported P values are 2-sided, with P values less than .05 considered to represent statistically significant results. Continuous data that were normally distributed were summarized as mean and standard deviation. Data that were not normally distributed were summarized as median and interquartile range (IQR). Categorical data were presented as frequency and percentage. Data completeness was extremely good; therefore, a complete case analysis was used, with less than 0.5% of patients excluded from the final model owing to missing data on risk factors. Statistical analyses were completed in June 2017. All analyses were performed using Stata/SE statistical software version 14.0 (StataCorp LP).

    Results
    Study Cohort

    Over the 5-year study period (April 1, 2009, to March 31, 2014), there were 152 383 index critical care admissions for sepsis and 98 732 sepsis survivors. After exclusions for age younger than 16 years (976 patients) and those sepsis survivors who were discharged from the hospital after March 31, 2014 (3008 patients), the sepsis survivor study cohort included 94 748 patients (Figure 1; eFigure in the Supplement).

    Cohort Characteristics and Long-term Mortality

    Overall, these sepsis survivors had a mean (SD) age of 61.3 (17.0) years, 43 584 (46.0%) were female, and 86 056 (90.8%) were white. Sepsis survivors more often had nonsurgical admissions (69.3%). The most common site of infection was respiratory (46.3%), the critical care admission day mean (SD) total APACHE II score was 16.7 (5.9) points, and the median (IQR) critical care length of stay was 4 (2-8) days. Further generic and sepsis-specific patient characteristics are reported in the Table. By 1 year after hospital discharge, 15% of sepsis survivors had died, with 6% to 8% dying per year over the next 5 years (Figure 2). Cumulative time-to-mortality plots indicated significant within-strata differences (Figure 3).

    Factors Independently Associated With Long-term Mortality

    Among the generic patient characteristics assessed, increasing age, being male, having 1 or more severe comorbidities, having some or total dependency prior to index critical care admission, and longer index hospital length of stay increased the risk of long-term mortality in sepsis survivors, consistent with published literature. Compared with sepsis survivors who survived their index critical care admission for sepsis between April 2009 and March 2010, sepsis survivors in later years had higher risk of long-term mortality (adjusted HR, 1.02; 95% CI, 1.01-1.03, equating to 2% increase in hazard per year; P < .001). Increasing score on the APACHE II acute physiology component was significantly associated with increased risk of long-term mortality (adjusted HR, 1.11 for every 5 points; 95% CI, 1.08-1.13; P < .001). Patients with surgical status had a lower risk of long-term mortality than those with nonsurgical (medical) status (Figure 4; eTable 2 in the Supplement).

    Among the sepsis-specific characteristics assessed, when compared with respiratory site of infection, gastrointestinal, genitourinary, musculoskeletal, and neurological sites of infection had adjusted HRs for long-term mortality less than 1. Compared with single organ dysfunction, having either 2 or 3 organ dysfunctions was associated with greater mortality (adjusted HR, 1.07; 95% CI, 1.01-1.13; and adjusted HR, 1.18; 95% CI, 1.03-1.14, respectively), while having 4 or more organ dysfunctions was not. Septic shock at index admission had an adjusted HR of 0.89 (95% CI, 0.85-0.92; P < .001) for long-term mortality in sepsis survivors (Figure 4; eTable 2 in the Supplement).

    Factors That Modify the Association of Acute Severity of Illness With Long-term Mortality

    The association of APACHE II acute physiology component score on longer-term mortality decreased by 4% per year for sepsis survivors with 1 comorbidity and 7% per year for those with 2 or more severe comorbidities. We observed a 4% increase in time-varying hazard per year for age. The time-varying reduction in the hazard for the APACHE II acute physiology component score was less than 2% per year, compared with the 9% decrease in hazard per year for 1 severe comorbidity and 20% decrease in hazard per year for 2 or more severe comorbidities, implying patients with more comorbidities died earlier during follow-up (Figure 4; eTable 2 in the Supplement).

    Sensitivity Analyses

    Results from sensitivity analyses were consistent with the primary analyses for the generic and sepsis-specific risk factors assessed. For example, all sensitivity analyses had similar adjusted HRs for increments in the APACHE II acute physiology component score (1) in sepsis survivors without severe comorbidities (n = 80 139; adjusted HR, 1.11 for every 5 points; 95% CI, 1.08-1.13), (2) from 117 critical care units that contributed data over the entire study period (n = 68 703; adjusted HR, 1.10 for every 5 points; 95% CI, 1.07-1.13), (3) for overlapping survival-time intervals, and (4) for random variation with shared frailty model (adjusted HR, 1.08 for every 5 points; 95% CI, 1.07-1.10) (eTable 2 in the Supplement). The relative survival analysis was comparable to previous reports, with survival decreasing from 87.3% by 1 year to 69.2% by 5 years (eTable 3 in the Supplement).

    Discussion

    In this cohort of adult sepsis survivors, as defined by Sepsis-3 criteria, both generic and sepsis-specific patient characteristics known during the index critical care admission for sepsis were independently associated with long-term mortality. Increasing age, male sex, 1 or more severe comorbidities, prehospitalization dependency, and nonsurgical (medical) status increased the risk of long-term mortality. The risk of long-term mortality differed by site of infection at index critical care admission for sepsis. Novel findings were the incremental increase in risk of long-term death in sepsis survivors associated with increased APACHE II acute physiology component score (adjusted HR, 1.11; 95% CI, 1.08-1.13 for every 5 points), lower risk associated with septic shock (adjusted HR, 0.89; 95% CI, 0.85-0.92), and the variation in additional hazard for long-term death associated with number of organ dysfunctions. The greater number of in-hospital deaths in patients with septic shock and in patients with greater organ dysfunction is the most likely explanation for these results,3,11 an important finding that is often overlooked when estimated using cumulative mortality.

    One of the key reasons why we studied a sepsis survivor population was to unmask the important risk factors associated with long-term mortality, which are often disguised by the risk factors associated with short-term mortality, such as acute illness severity. The incremental increase in hazard seen with increasing APACHE II acute physiology component score persisted in the sensitivity analysis, excluding patients with chronic severe comorbidity. This finding, when considered in the context of clinical11 and biological22 differences within sepsis cohorts, reinforces the value of how routinely available clinical information on index critical care admission for sepsis could inform follow-up care of sepsis survivors. This reasoning is strengthened when noting the differences in the impact of site of infection on long-term mortality we observed, as site of infection potentially influences immune responses.23 A related argument in this context is that even in 2 patients with similar predicted risk but with differences in relative contributions of generic and sepsis-specific risk components, differences would exist in treatability of long-term sequalae. Thus, when these factors independently associated with long-term mortality from our results are considered with those from the literature,5,6,8,10,24,25 it indicates that sepsis survivors both retain risk from index critical care admission for sepsis but also accumulate additional risk following hospital discharge from their worsening preexisting comorbidities26 and/or new comorbidities (eg, worsening cognition,24 cardiovascular morbidity27), both of which are associated with increased risk of long-term mortality. Thus, our study makes a case for understanding how management following discharge could alter sepsis survivors’ risk of long-term adverse outcomes.26

    Strengths

    We used the Case Mix Programme database, which is a national database for critical care units in England and has been independently assessed to be of high quality.14 We reduced selection bias by studying consecutive critical care admissions for sepsis from 192 critical care units representing the whole critical care population in a country over a 6-year period and commencing patient follow-up from hospital discharge with 96% data linkage for long-term mortality. Long-term mortality from a single health care system is often challenged on external validity owing to differences in clinical care.28 We have maximized the external validity of our study by using a valid Sepsis-3 sepsis case definition in a nationally representative adult sepsis survivor cohort older than 16 years. These are important strengths, as, of the 2 recent long-term sepsis outcome studies, 1 focused on older adults (≥65 years) with insurance coverage admitted between 2002 and 2010,8 and the second reported sepsis cases between 2000 and 2002 from a national health insurance database.7

    Limitations

    This study has limitations. As seen in all follow-up studies, our cohort had different durations of follow-up. We used proportional hazards modeling to account for differences in follow-up duration and adjusted for common confounders6,10 in the association between sepsis and long-term mortality based on systematic review of literature.6 Our study cohort reports the epidemiology of adult sepsis survivors, who all had critical care admissions with sepsis identified in the first 24 hours of admission. Thus, we will underestimate patients with nonsepsis critical care admissions who develop sepsis later in their critical care stay. However, as England has among the lowest per capita critical care bed provision in Europe (3.5-7.4 per 100 000 population),29 probability of underestimation is low, as organ dysfunction will often be present on admission day. We were not able to study the biological mechanisms that could explain our findings. Information bias due to misclassification of sepsis exposure or long-term mortality is a risk that we reduced by deriving sepsis cases using raw physiology and long-term mortality from a high-quality national data source. As our goal was to identify risk factors, we have only reported age- and sex- matched general population controls for comparison with those studies that quantify excess long-term risk of death from sepsis.7-9 A marker of physiological disturbance closer to discharge would be potentially useful, with the counterpoints that the median critical care length of stay was 4 days and if there was significant physiological disturbance, patients would not be discharged from the hospital. We do not present a risk score to implement at bedside, as this was not our research question. Although we did not estimate a formal sample size calculation, we do report the long-term mortality from the largest adult sepsis survivor cohort to date.

    Our observations that acute illness severity and site of infection at index admission are associated with increased risk of long-term mortality in sepsis survivors has biological plausibility and informs future research. Acute illness severity and illness trajectory seen with critical care admissions for sepsis are associated with health before the illness causing hospitalization, immune responses in sepsis,22,30,31 and early appropriate clinical management.32-34 Immune abnormalities seen in patients with sepsis31 appear to persist in sepsis survivors.35,36 New or relapsed infection is one of the common reasons for rehospitalization in sepsis survivors, which is potentially associated with site of infection at index sepsis admission.10,37 Thus, observational studies identifying biological characteristics and value of parsimonious clinical prediction tools using generic and sepsis-specific risk factors to predict risk of long-term adverse health in sepsis survivors would be valuable. The future interventional trials may include clinical practice interventions such as extended follow-up care either by primary care physicians5 or with targeted follow-up clinics38 or immunological interventions informed by acute sepsis illness biology.39,31,40 Our study adds further impetus to the recently adopted World Health Organization resolution calling for further work informing “policy decisions related to preventive, diagnostic and treatment activities and access to relevant health care for sepsis survivors.”41

    Conclusions

    Generic and sepsis-specific risk factors, known during index critical care admission for sepsis, are important risk factors associated with long-term mortality seen in a sepsis survivor cohort. Our research provides validity to target sepsis survivor populations based on index admission characteristics, for biological characterization and designing interventions to reduce long-term mortality.

    Back to top
    Article Information

    Accepted for Publication: April 13, 2019.

    Published: May 31, 2019. doi:10.1001/jamanetworkopen.2019.4900

    Correction: This article was corrected on June 21, 2019, to fix an incorrect academic degree in the byline.

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Shankar-Hari M et al. JAMA Network Open.

    Corresponding Author: Manu Shankar-Hari, MSc, PhD, FRCA, FFICM, ICU Critical Care Offices, St Thomas’ Hospital, Guy’s and St Thomas’ NHS Foundation Trust, London SE17EH, United Kingdom (manu.shankar-hari@kcl.ac.uk).

    Author Contributions: Drs Shankar-Hari and Harrison had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Shankar-Hari, Harrison, Rubenfeld, Rowan.

    Acquisition, analysis, or interpretation of data: Shankar-Hari, Harrison, Ferrando-Vivas, Rowan.

    Drafting of the manuscript: Shankar-Hari, Rubenfeld, Rowan.

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

    Statistical analysis: Shankar-Hari, Harrison, Ferrando-Vivas.

    Obtained funding: Shankar-Hari.

    Administrative, technical, or material support: Rowan.

    Supervision: Harrison, Rubenfeld, Rowan.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: Dr Shankar-Hari is supported by National Institute for Health Research Clinician Scientist Award CS-2016-16-011.

    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.

    Disclaimer: The views expressed in this publication are those of the authors and not necessarily those of the National Health Service, the National Institute for Health Research, or the Department of Health and Social Care. Dr Rubenfeld, a JAMA Network Open associate editor, was not involved in the editorial review of or the decision to publish this article.

    Additional Information: The data used in this study are accessible on request from the UK national intensive care admissions data set managed by the Intensive Care National Audit and Research Centre.

    References
    1.
    Singer  M, Deutschman  CS, Seymour  CW,  et al.  The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).  JAMA. 2016;315(8):801-810. doi:10.1001/jama.2016.0287PubMedGoogle ScholarCrossref
    2.
    Fleischmann  C, Scherag  A, Adhikari  NK,  et al; International Forum of Acute Care Trialists.  Assessment of global incidence and mortality of hospital-treated sepsis: current estimates and limitations.  Am J Respir Crit Care Med. 2016;193(3):259-272. doi:10.1164/rccm.201504-0781OCPubMedGoogle ScholarCrossref
    3.
    Shankar-Hari  M, Harrison  DA, Rubenfeld  GD, Rowan  K.  Epidemiology of sepsis and septic shock in critical care units: comparison between sepsis-2 and sepsis-3 populations using a national critical care database.  Br J Anaesth. 2017;119(4):626-636. doi:10.1093/bja/aex234PubMedGoogle ScholarCrossref
    4.
    Kaukonen  KM, Bailey  M, Suzuki  S, Pilcher  D, Bellomo  R.  Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000-2012.  JAMA. 2014;311(13):1308-1316. doi:10.1001/jama.2014.2637PubMedGoogle ScholarCrossref
    5.
    Prescott  HC, Angus  DC.  Enhancing recovery from sepsis: a review.  JAMA. 2018;319(1):62-75. doi:10.1001/jama.2017.17687PubMedGoogle ScholarCrossref
    6.
    Shankar-Hari  M, Ambler  M, Mahalingasivam  V, Jones  A, Rowan  K, Rubenfeld  GD.  Evidence for a causal link between sepsis and long-term mortality: a systematic review of epidemiologic studies.  Crit Care. 2016;20(1):101. doi:10.1186/s13054-016-1276-7PubMedGoogle ScholarCrossref
    7.
    Ou  SM, Chu  H, Chao  PW,  et al.  Long-term mortality and major adverse cardiovascular events in sepsis survivors. a nationwide population-based study.  Am J Respir Crit Care Med. 2016;194(2):209-217. doi:10.1164/rccm.201510-2023OCPubMedGoogle ScholarCrossref
    8.
    Prescott  HC, Osterholzer  JJ, Langa  KM, Angus  DC, Iwashyna  TJ.  Late mortality after sepsis: propensity matched cohort study.  BMJ. 2016;353:i2375. doi:10.1136/bmj.i2375PubMedGoogle ScholarCrossref
    9.
    Davis  JS, He  V, Anstey  NM, Condon  JR.  Long term outcomes following hospital admission for sepsis using relative survival analysis: a prospective cohort study of 1,092 patients with 5 year follow up.  PLoS One. 2014;9(12):e112224. doi:10.1371/journal.pone.0112224PubMedGoogle ScholarCrossref
    10.
    Shankar-Hari  M, Rubenfeld  GD.  Understanding long-term outcomes following sepsis: implications and challenges.  Curr Infect Dis Rep. 2016;18(11):37. doi:10.1007/s11908-016-0544-7PubMedGoogle ScholarCrossref
    11.
    Shankar-Hari  M, Harrison  DA, Rowan  KM.  Differences in impact of definitional elements on mortality precludes international comparisons of sepsis epidemiology-a cohort study illustrating the need for standardized reporting.  Crit Care Med. 2016;44(12):2223-2230. doi:10.1097/CCM.0000000000001876PubMedGoogle ScholarCrossref
    12.
    Vandenbroucke  JP, von Elm  E, Altman  DG,  et al; STROBE Initiative.  Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration.  PLoS Med. 2007;4(10):e297. doi:10.1371/journal.pmed.0040297PubMedGoogle ScholarCrossref
    13.
    Young  JD, Goldfrad  C, Rowan  K.  Development and testing of a hierarchical method to code the reason for admission to intensive care units: the ICNARC Coding Method.  Br J Anaesth. 2001;87(4):543-548. doi:10.1093/bja/87.4.543PubMedGoogle ScholarCrossref
    14.
    Harrison  DA, Brady  AR, Rowan  K.  Case mix, outcome and length of stay for admissions to adult, general critical care units in England, Wales and Northern Ireland: the Intensive Care National Audit & Research Centre Case Mix Programme Database.  Crit Care. 2004;8(2):R99-R111. doi:10.1186/cc2834PubMedGoogle ScholarCrossref
    15.
    NHS Digital. Linked mortality data from Hospital Episode Statistics and the Office for National Statistics. http://content.digital.nhs.uk/hesonsmortality. Accessed April 4, 2017.
    16.
    Shankar-Hari  M, Phillips  GS, Levy  ML,  et al; Sepsis Definitions Task Force.  Developing a new definition and assessing new clinical criteria for septic shock: for the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).  JAMA. 2016;315(8):775-787. doi:10.1001/jama.2016.0289PubMedGoogle ScholarCrossref
    17.
    Knaus  WA, Draper  EA, Wagner  DP, Zimmerman  JE.  APACHE II: a severity of disease classification system.  Crit Care Med. 1985;13(10):818-829. doi:10.1097/00003246-198510000-00009PubMedGoogle ScholarCrossref
    18.
    Grambsch  PM, Therneau  TM.  Proportional hazards tests and diagnostics based on weighted residuals.  Biometrika. 1994;81(3):515-526. doi:10.1093/biomet/81.3.515Google ScholarCrossref
    19.
    Austin  PC.  A tutorial on multilevel survival analysis: methods, models and applications.  Int Stat Rev. 2017;85(2):185-203. doi:10.1111/insr.12214PubMedGoogle ScholarCrossref
    20.
    Dickman  PW, Coviello  E.  Estimating and modeling relative survival.  Stata J. 2015;15(1):186-215. doi:10.1177/1536867X1501500112Google ScholarCrossref
    21.
    Office for National Statistics. https://www.ons.gov.uk. Accessed February 28, 2017.
    22.
    Scicluna  BP, van Vught  LA, Zwinderman  AH,  et al; MARS consortium.  Classification of patients with sepsis according to blood genomic endotype: a prospective cohort study.  Lancet Respir Med. 2017;5(10):816-826. doi:10.1016/S2213-2600(17)30294-1PubMedGoogle ScholarCrossref
    23.
    Burnham  KL, Davenport  EE, Radhakrishnan  J,  et al.  Shared and distinct aspects of the sepsis transcriptomic response to fecal peritonitis and pneumonia.  Am J Respir Crit Care Med. 2017;196(3):328-339. doi:10.1164/rccm.201608-1685OCPubMedGoogle ScholarCrossref
    24.
    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
    25.
    Yende  S, D’Angelo  G, Kellum  JA,  et al; GenIMS Investigators.  Inflammatory markers at hospital discharge predict subsequent mortality after pneumonia and sepsis.  Am J Respir Crit Care Med. 2008;177(11):1242-1247. doi:10.1164/rccm.200712-1777OCPubMedGoogle ScholarCrossref
    26.
    Prescott  HC, Langa  KM, Iwashyna  TJ.  Readmission diagnoses after hospitalization for severe sepsis and other acute medical conditions.  JAMA. 2015;313(10):1055-1057. doi:10.1001/jama.2015.1410PubMedGoogle ScholarCrossref
    27.
    Corrales-Medina  VF, Alvarez  KN, Weissfeld  LA,  et al.  Association between hospitalization for pneumonia and subsequent risk of cardiovascular disease.  JAMA. 2015;313(3):264-274. doi:10.1001/jama.2014.18229PubMedGoogle ScholarCrossref
    28.
    Wunsch  H, Angus  DC, Harrison  DA, Linde-Zwirble  WT, Rowan  KM.  Comparison of medical admissions to intensive care units in the United States and United Kingdom.  Am J Respir Crit Care Med. 2011;183(12):1666-1673. doi:10.1164/rccm.201012-1961OCPubMedGoogle ScholarCrossref
    29.
    Rhodes  A, Ferdinande  P, Flaatten  H, Guidet  B, Metnitz  PG, Moreno  RP.  The variability of critical care bed numbers in Europe.  Intensive Care Med. 2012;38(10):1647-1653. doi:10.1007/s00134-012-2627-8PubMedGoogle ScholarCrossref
    30.
    Kellum  JA, Kong  L, Fink  MP,  et al; GenIMS Investigators.  Understanding the inflammatory cytokine response in pneumonia and sepsis: results of the Genetic and Inflammatory Markers of Sepsis (GenIMS) study.  Arch Intern Med. 2007;167(15):1655-1663. doi:10.1001/archinte.167.15.1655PubMedGoogle ScholarCrossref
    31.
    van der Poll  T, van de Veerdonk  FL, Scicluna  BP, Netea  MG.  The immunopathology of sepsis and potential therapeutic targets.  Nat Rev Immunol. 2017;17(7):407-420. doi:10.1038/nri.2017.36PubMedGoogle ScholarCrossref
    32.
    Rhodes  A, Evans  LE, Alhazzani  W,  et al.  Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016.  Intensive Care Med. 2017;43(3):304-377. doi:10.1007/s00134-017-4683-6PubMedGoogle ScholarCrossref
    33.
    Seymour  CW, Gesten  F, Prescott  HC,  et al.  Time to treatment and mortality during mandated emergency care for sepsis.  N Engl J Med. 2017;376(23):2235-2244. doi:10.1056/NEJMoa1703058PubMedGoogle ScholarCrossref
    34.
    Liu  VX, Fielding-Singh  V, Greene  JD,  et al.  The timing of early antibiotics and hospital mortality in sepsis.  Am J Respir Crit Care Med. 2017;196(7):856-863. doi:10.1164/rccm.201609-1848OCPubMedGoogle ScholarCrossref
    35.
    Riché  F, Chousterman  BG, Valleur  P, Mebazaa  A, Launay  JM, Gayat  E.  Protracted immune disorders at one year after ICU discharge in patients with septic shock.  Crit Care. 2018;22(1):42. doi:10.1186/s13054-017-1934-4PubMedGoogle ScholarCrossref
    36.
    Arens  C, Bajwa  SA, Koch  C,  et al.  Sepsis-induced long-term immune paralysis—results of a descriptive, explorative study.  Crit Care. 2016;20(1):93. doi:10.1186/s13054-016-1233-5PubMedGoogle ScholarCrossref
    37.
    DeMerle  KM, Royer  SC, Mikkelsen  ME, Prescott  HC.  Readmissions for recurrent sepsis: new or relapsed infection?  Crit Care Med. 2017;45(10):1702-1708. doi:10.1097/CCM.0000000000002626PubMedGoogle ScholarCrossref
    38.
    Shen  E, Koyama  SY, Huynh  DN,  et al.  Association of a dedicated post-hospital discharge follow-up visit and 30-day readmission risk in a Medicare Advantage population.  JAMA Intern Med. 2017;177(1):132-135. doi:10.1001/jamainternmed.2016.7061PubMedGoogle ScholarCrossref
    40.
    Shankar-Hari  M, Fear  D, Lavender  P,  et al.  Activation-associated accelerated apoptosis of memory B cells in critically ill patients with sepsis.  Crit Care Med. 2017;45(5):875-882. doi:10.1097/CCM.0000000000002380PubMedGoogle ScholarCrossref
    41.
    Reinhart  K, Daniels  R, Kissoon  N, Machado  FR, Schachter  RD, Finfer  S.  Recognizing sepsis as a global health priority—a WHO resolution.  N Engl J Med. 2017;377(5):414-417. doi:10.1056/NEJMp1707170PubMedGoogle ScholarCrossref
    ×