Failure to recover to baseline HRQL at follow-up occurred in 23.8% of patients (n = 188), while 34.9% (n = 276) recovered to baseline and 41.5% (n = 328) improved to greater than or equal to 4.5 PedsQL points above baseline. Patients with HRQL improvement were more likely to have chronic illnesses and be immunosuppressed compared with patients who recovered within 4.5 points of baseline; they also had longer time to follow-up (Table S1, Supplemental Digital Content 1, http://links.lww.com/PCC/A908). A similar proportion of patients without sepsis who were enrolled in the OAP failed to recover (24.1%), but with earlier median follow-up than the sepsis cohort. By 60 days (±10) from discharge, 20.4% of sepsis patients had failure to recover versus 16.4% of nonsepsis patients, whereas 52.0% of sepsis patients and 59.0% of nonsepsis patients had HRQL improvement.
The median PedsQL scores for the entire sepsis cohort were 88.5 (interquartile range [IQR], 76.4–95.3) at baseline, 51.9 (35.9–71.6) at admission, and 93.6 (79.8–98.7) at follow-up (Fig. 2). The median baseline score of sepsis patients who failed to recover to their baseline status after discharge was similar to the overall sepsis cohort (90.6; IQR, 80.7–96.1), but fell further at admission to a median score of 42.8 (30.2–61.4) with a median follow-up score of 71.1 (59.8–82.2), a decline of 19.5 points from baseline.
Among the entire sepsis cohort, there was no significant change from baseline to follow-up scores on either the physical or psychosocial subscales (Fig. 3). The group of patients who recovered to baseline experienced median increases of 3.1 points on the physical subscale (IQR, 0.0–12.5) and 6.9 points on the psychosocial subscale (0.0–16.7). The group of patients who failed to recover experienced a median physical subscale decline of 18.8 points (−34.4 to 9.4) and a median psychosocial subscale decline of 10.2 points (−20.0 to 3.3).
We evaluated factors associated with failure to recover by comparing patients with sepsis who did and did not recover across four categories of potential risk factors: demographics, medical history, infection characteristics, and illness severity (Table 2). Demographic factors associated with failure to recover included older age and private insurance. Age greater than or equal to 12 years was associated most strongly with failure to recover compared with other age categories, with 30.1% failing to recover to baseline. Older patients were more likely to have complex chronic conditions, immune compromise, and septic shock (p < 0.001). Patients with private insurance were older than those with public insurance (p = 0.005), but did not differ on other measures.
Components of medical history associated with failure to recover included chronic condition category and immune status. One-third of patients with complex chronic disease did not recover to their baseline HRQL compared with 23.1% of those with no chronic disease and 16.7% of those with noncomplex chronic disease (p = 0.005). Failure to recover occurred in 40.7% of immunocompromised patients versus 21.0% of nonimmunocompromised patients (p < 0.001). Patients with complex chronic disease were more likely to be immunocompromised (p < 0.001); neither chronic condition category nor immune status was associated with sepsis category.
Source of infection was the only infection characteristic associated with outcome, with failure to recover occurring in 58.6% of patients with primary bacteremia and 47.4% of patients with CNS infections (p < 0.001). Neither the presence of a positive culture nor whether the sepsis was bacterial, fungal, or viral conferred increased risk.
The category of sepsis at admission was strongly associated with outcome. Failure to recover to baseline HRQL occurred in 53.7% of patients with septic shock, 29.4% of patients with severe sepsis without shock, and 21.8% of patients with sepsis alone (p < 0.001). ICU admission was also associated with failure to recover, with 38.5% of ICU patients failing to recover versus 21.9% of those without an ICU stay (p < 0.001). Finally, longer hospital length of stay (LOS) was associated with failure to recover, with a threshold of greater than or equal to 7 days most strongly associated; 44.4% of those with LOS greater than or equal to 7 days failed to return to baseline compared with 20.5% with LOS less than 7 days (p < 0.001).
Time to follow-up ranged from 14 to 137 days after hospital discharge (median, 31 d; IQR, 21–47). The proportion of patients who failed to recover decreased over time up to 8 weeks post-discharge, but this trend did not continue among patients with later follow-up assessments (Fig. 4). Shorter time to follow-up was associated with a higher likelihood of failure to recover on bivariate analysis.
After controlling for time to follow-up, multivariable analysis demonstrated that older age, immune compromise, septic shock, and longer hospital LOS were independently associated with increased risk of failure to recover to baseline (Table 3). Immune compromise conferred the highest relative risk of 1.83 compared to those without immune compromise (95% CI, 1.40–2.40), followed by septic shock with a relative risk of 1.79 compared to sepsis (95% CI, 1.24–2.58). Risk of failure to recover increased by 2% per year of age (95% CI, 1.01–1.05), and by 3% with each additional hospital day (95% CI, 1.01–1.04). Time to follow-up was no longer significantly associated with failure to recover on multivariable analysis. Sensitivity analysis using a comparison group of only patients who recovered within 4.5 points of baseline demonstrated similar bivariate and multivariable results (Table S2, Supplemental Digital Content 2, http://links.lww.com/PCC/A909).
Decrease in PedsQL score from baseline to follow-up among ward patients was associated with immune compromise and longer hospital LOS (Table S3, Supplemental Digital Content 3, http://links.lww.com/PCC/A910). Among ICU patients, a decrease in score was associated with older age, immune compromise, soft-tissue infection, and longer hospital LOS (Table S4, Supplemental Digital Content 4, http://links.lww.com/PCC/A911).
In this study, we provide the first description of the trajectory of children’s HRQL status before, during, and after a hospitalization for sepsis. Previous studies have demonstrated lower HRQL among PICU survivors (8, 14, 15) and survivors of meningococcemia (16) and meningitis (17) compared with population norms, but there are no data evaluating HRQL after pediatric sepsis in general, or comparing individual patient HRQL outcomes to their own baseline status. The majority of patients in this 4-year cohort of children hospitalized with community-acquired sepsis fully recovered to their baseline status after hospital discharge or even improved, consistent with prior studies of ill and injured children (30, 36) and patients without sepsis at our institution. Nearly one-quarter of patients, however, experienced a clinically significant decline in HRQL from their pre-admission baseline to their post-discharge follow-up. The median persistent declination in this group was 19.5 points below baseline, representing a decline of over four times the minimum clinically significant difference and a 22% decline from the median baseline score. Evaluating how this group differs from the patients who recovered fully or improved may help identify targets for interventions to enhance post-discharge outcomes for at-risk children.
We identified multiple factors associated with HRQL deterioration spanning all four domains of focus: demographics, pre-existing health status, infection characteristics, and illness severity. Older age was the demographic factor most strongly associated with failure to recover. Older age has previously been demonstrated to be a risk factor for poor HRQL compared with population norms among PICU patients (15). This may suggest greater resilience among younger children recovering from illness, less morbidity from illnesses requiring hospitalization for younger patients (e.g., bronchiolitis), or that rapidly developing infants and toddlers may manifest impact on HRQL by lack of developmentally expected improvement rather than decline.
Chronic comorbidities were associated with failure to recover only on bivariate analysis and were in fact associated with slight improvements in follow-up score in multivariable linear regression models. Previous studies have demonstrated that patients with chronic health conditions are at risk for poor outcomes after critical illness (2, 4, 9, 15, 37). In our cohort, many patients with chronic comorbidities were also immunocompromised from cancer, rheumatologic conditions, or organ transplants. Immune compromise was strongly associated with poor outcome across all analyses. When controlling for immunocompromised status, the remaining chronic health conditions no longer conferred excess risk for HRQL decline, suggesting that conditions associated with immune compromise may place patients at highest risk of adverse outcomes. Notably, however, patients with chronic illness and immune compromise were also more likely to experience HRQL improvement, perhaps due to better control of underlying conditions among these patients while hospitalized.
Our finding that bacterial infection was not associated with increased risk of adverse outcomes highlights the impact of viral and culture-negative sepsis, with nearly one-quarter of these patients failing to recover to baseline. Existing literature in adult populations on the association of culture-negative sepsis with outcomes is mixed (38, 39). When a specific infection was identified in our cohort, however, the source did matter; prevalence of failure to recover ranged from under 20% for soft-tissue infections to nearly 50% for CNS infections and nearly 60% for bacteremia. The range of outcomes even within these subcategories is apparent when stratifying by ICU status, where soft-tissue infections in the ICU population were associated with an average decline of 17 points between baseline and follow-up, presumably reflecting the impact of severe mucocutaneous and necrotizing soft-tissue infections.
PRISM III, PELOD, and MPEWS scores were not associated with outcome, likely reflecting the low illness severity in our cohort. When evaluating illness severity by sepsis category, however, severe sepsis and septic shock were highly correlated with failure to recover to baseline. Other indicators of illness severity such as ICU admission and hospital LOS were also associated with adverse outcome; longer LOS was, in fact, the only factor other than immune compromise that was associated with failure to recover across all analyses.
Although most of these factors are not modifiable, they do allow identification of patients at highest risk for poor HRQL outcome. Using the linear regression modeling results, we can estimate how a given patient with multiple risk factors may do after discharge. For example, a previously healthy 12-year-old admitted to the ICU with meningitis requiring a 20-day hospital stay would be estimated to have a decline in PedsQL score by 38.3 points (95% CI, 21.8–54.9) from baseline at 28 days after discharge.
These findings also support the use of HRQL as an outcome measure following sepsis, distinguishing between patients with varying levels of sepsis severity, chronic conditions, and prolonged hospitalizations. Mortality has been the most common outcome measure used in interventional sepsis trials over the past 30 years, but as mortality continues to decline we are in need of an outcome measure that is clinically meaningful and captures additional patient-centered effects. Historically, pediatric critical care interventional trials have not reported HRQL as an outcome measure, and there has been no consensus on when or how to measure HRQL or data describing how clinical factors or interventions may affect HRQL (40).
We have found that persistent declines in HRQL occur in a large proportion of patients surviving sepsis, suggesting that it may be a more sensitive outcome measure than mortality to evaluate differences between groups and epidemiologic trends. Failure to recover to baseline HRQL may also be a more patient-centered and clinically meaningful outcome measure than physiologic parameters such as duration of shock or mechanical ventilation (9, 40).
Our data highlight the importance of comparing HRQL outcomes to patients’ own baseline status rather than to population norms as has previously been reported. While the proportion of patients with a follow-up PedsQL score greater than or equal to 4.5 points below the population mean of 84.1 points (24.9%) was similar to the proportion of patients with a follow-up score greater than or equal to 4.5 points below their individual baseline score (23.8%), there was only moderate overlap between the sets of patients identified by these two methods. Thirty-four percent of the patients who were below the population mean at follow-up were not significantly below their individual baseline score, and only 69% of the patients who had a persistent clinically significant decline from their individual baseline score would have been identified as being below the population norm.
Finally, HRQL assessment is also applicable to patients across the range of illness severity. Previous evaluations of outcomes among children surviving sepsis have all focused on patients with severe sepsis or septic shock. To our knowledge, there are no data evaluating outcomes in non-ICU patients. The responsiveness, wide applicability, patient-centeredness, and potential for improved sensitivity all support the implementation of HRQL outcome assessment in future sepsis research.
There were several limitations to this study. Only a subset of inpatients was sampled for HRQL assessment, and our cohort had lower illness severity and fewer comorbidities than the general population of patients with sepsis admitted to our institution. Reluctance of research assistants to approach families of the sickest patients and lower consent rates from these families likely contributed to difficulty enrolling the most acutely ill patients. This incomplete sampling may limit generalizability to more severely ill populations, though most likely biases our results toward the null given the association between higher illness severity and worse HRQL outcomes.
An additional limitation was that our assessment of patients’ baseline HRQL status was based on recall, which is inherently subject to bias. The direction of recall bias in this situation is unknown and likely variable; some families may recall their child to have had a better prior HRQL in comparison to the acute illness than they actually did, whereas others may be influenced by the acute illness or a potential prodrome such that they recall their child’s baseline status to be worse than it actually was. The finding that many patients had an improvement in HRQL following discharge supports the possibility that recall was biased toward lower HRQL, though improvement could also be due to normal variation, improvement in HRQL after treatment for an underlying condition, or developmental progression of physical, social, and cognitive functioning. Although several studies have demonstrated construct validity of the PedsQL for evaluating baseline HRQL based on recall (28–30), this has not been prospectively studied and a tendency toward positive recall of baseline HRQL in our cohort would bias our results toward a higher proportion of patients with failure to recover.
A final limitation was that follow-up was performed at variable time-points with overrepresentation of earlier time-points, introducing the potential for sampling-time bias and preventing direct comparison across patients at equivalent intervals between discharge and follow-up. Our data suggest an improvement in HRQL over time, though this was not statistically significant on multivariable analysis. Further research is needed to better evaluate how HRQL varies over time in order to more precisely describe its use as an outcome measure and to determine which factors contribute to persistent HRQL declines.
Despite these limitations, this study contributes substantially to our understanding of the ongoing burden of sepsis experienced by both ICU and non-ICU patients after hospital discharge. Our cohort of 790 patients represents one of the largest studies of HRQL outcomes in hospitalized children and to our knowledge is the largest evaluation of morbidity following pediatric sepsis. This study complements the ongoing multicenter Life After Pediatric Sepsis Evaluation (R01HD073362) by assessing outcomes across the spectrum of sepsis severity including non-ICU patients.
HRQL impairment is common among children surviving sepsis, with nearly one-quarter of children experiencing a clinically significant deterioration in HRQL from their baseline status at the time of post-discharge follow-up. Failure to recover to baseline HRQL is associated with illness severity but importantly also occurs frequently in children with sepsis who do not require ICU care, a population for which outcomes have not previously been characterized. Identification of factors associated with HRQL deterioration is essential to guide interventions to improve long-term outcomes for at-risk children. Our data identify risk factors for adverse outcomes following sepsis and imply the need for ongoing evaluation and treatment by outpatient providers for all survivors of pediatric sepsis for at least the first 3 months after hospital discharge, with particular attention to teenagers, those who hospitalized for more than a week, immunocompromised patients, and those who experienced septic shock.
1. Balamuth F, Weiss SL, Neuman MI, et al. Pediatric severe sepsis
in U.S. children’s hospitals. Pediatr Crit Care Med 2014; 15:798–805
2. Czaja AS, Zimmerman JJ, Nathens AB. Readmission and late mortality after pediatric severe sepsis
. Pediatrics 2009; 123:849–857
3. Hartman ME, Linde-Zwirble WT, Angus DC, et al. Trends in the epidemiology of pediatric severe sepsis
*. Pediatr Crit Care Med 2013; 14:686–693
4. Ruth A, McCracken CE, Fortenberry JD, et al. Pediatric severe sepsis
: Current trends and outcomes from the Pediatric Health Information Systems database. Pediatr Crit Care Med 2014; 15:828–838
5. de Souza DC, Barreira ER, Faria LS. The epidemiology of sepsis
in childhood. Shock 2017; 47:2–5
6. Farris RW, Weiss NS, Zimmerman JJ. Functional outcomes in pediatric severe sepsis
: Further analysis of the researching severe sepsis
and organ dysfunction in children: A global perspective trial. Pediatr Crit Care Med 2013; 14:835–842
7. Weiss SL, Fitzgerald JC, Pappachan J, et al; Sepsis
Prevalence, Outcomes, and Therapies (SPROUT) Study Investigators and Pediatric Acute Lung Injury and Sepsis
Investigators (PALISI) Network: Global epidemiology of pediatric severe sepsis
: The sepsis
prevalence, outcomes, and therapies study. Am J Respir Crit Care Med 2015; 191:1147–1157
8. Conlon NP, Breatnach C, O’Hare BP, et al. Health-related quality of life
after prolonged pediatric intensive care unit stay. Pediatr Crit Care Med 2009; 10:41–44
9. Aspesberro F, Mangione-Smith R, Zimmerman JJ. Health-related quality of life
following pediatric critical illness. Intensive Care Med 2015; 41:1235–1246
11. Namachivayam P, Taylor A, Montague T, et al. Long-stay children in intensive care: Long-term functional outcome and quality of life
from a 20-yr institutional study. Pediatr Crit Care Med 2012; 13:520–528
12. Ebrahim S, Singh S, Hutchison JS, et al. Adaptive behavior, functional outcomes, and quality of life
outcomes of children requiring urgent ICU admission. Pediatr Crit Care Med 2013; 14:10–18
13. Polic B, Mestrovic J, Markic J, et al. Long-term quality of life
of patients treated in paediatric intensive care unit. Eur J Pediatr 2013; 172:85–90
14. Colville GA, Pierce CM. Children’s self-reported quality of life
after intensive care treatment. Pediatr Crit Care Med 2013; 14:e85–e92
15. Morrison AL, Gillis J, O’Connell AJ, et al. Quality of life
of survivors of pediatric intensive care. Pediatr Crit Care Med 2002; 3:1–5
16. Buysse CM, Raat H, Hazelzet JA, et al. Surviving meningococcal septic shock: Health consequences and quality of life
in children and their parents up to 2 years after pediatric intensive care unit discharge. Crit Care Med 2008; 36:596–602
17. Edmond K, Dieye Y, Griffiths UK, et al. Prospective cohort study of disabling sequelae and quality of life
in children with bacterial meningitis in urban Senegal. Pediatr Infect Dis J 2010; 29:1023–1029
18. Bronner MB, Knoester H, Sol JJ, et al. An explorative study on quality of life
and psychological and cognitive function in pediatric survivors of septic shock. Pediatr Crit Care Med 2009; 10:636–642
19. Varni JW, Seid M, Kurtin PS. PedsQL 4.0: Reliability and validity of the Pediatric Quality of Life
Inventory version 4.0 generic core scales in healthy and patient populations. Med Care 2001; 39:800–812
20. Goldstein B, Giroir B, Randolph A; International Consensus Conference on Pediatric Sepsis
: International pediatric sepsis
consensus conference: Definitions for sepsis
and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005; 6:2–8
21. Gebara BM. Values for systolic blood pressure. Pediatr Crit Care Med 2005; 6:500; author reply 500–500; author reply 501
22. Varni JW, Burwinkle TM, Seid M, et al. The PedsQL 4.0 as a pediatric population health measure: Feasibility, reliability, and validity. Ambul Pediatr 2003; 3:329–341
23. Varni JW, Limbers CA, Burwinkle TM. Parent proxy-report of their children’s health-related quality of life
: An analysis of 13,878 parents’ reliability and validity across age subgroups using the PedsQL 4.0 Generic Core Scales. Health Qual Life Outcomes 2007; 5:2
24. Varni JW, Limbers CA, Neighbors K, et al. The PedsQL™ Infant Scales: Feasibility, internal consistency reliability, and validity in healthy and ill infants. Qual Life Res 2011; 20:45–55
25. Limbers CA, Newman DA, Varni JW. Factorial invariance of child
self-report across race/ethnicity groups: A multigroup confirmatory factor analysis approach utilizing the PedsQL 4.0 Generic Core Scales. Ann Epidemiol 2009; 19:575–581
26. Desai AD, Zhou C, Stanford S, et al. Validity and responsiveness of the pediatric quality of life
inventory (PedsQL) 4.0 generic core scales in the pediatric inpatient setting. JAMA Pediatr 2014; 168:1114–1121
27. Aspesberro F, Fesinmeyer MD, Zhou C, et al. Construct validity and responsiveness of the Pediatric Quality of Life
Inventory 4.0 Generic Core Scales and Infant Scales in the PICU. Pediatr Crit Care Med 2016; 17:e272–e279
28. Kruse S, Schneeberg A, Brussoni M. Construct validity and impact of mode of administration of the PedsQL™ among a pediatric injury population. Health Qual Life Outcomes 2014; 12:168
29. Rivara FP, Vavilala MS, Durbin D, et al. Persistence of disability 24 to 36 months after pediatric traumatic brain injury: A cohort study. J Neurotrauma 2012; 29:2499–2504
30. Rabbitts JA, Palermo TM, Zhou C, et al. Pain and health-related quality of life
after pediatric inpatient surgery. J Pain 2015; 16:1334–1341
31. Virtual Pediatric Systems, LLC. Available at: www.myvps.org
. Accessed August 15, 2015
32. Simon TD, Cawthon ML, Stanford S, et al; Center of Excellence on Quality of Care Measures for Children with Complex Needs (COE4CCN) Medical Complexity Working Group: Pediatric medical complexity algorithm: A new method to stratify children by medical complexity. Pediatrics 2014; 133:e1647–e1654
33. Pollack MM, Patel KM, Ruttimann UE. PRISM III: An updated Pediatric Risk of Mortality score. Crit Care Med 1996; 24:743–752
34. Leteurtre S, Martinot A, Duhamel A, et al. Development of a pediatric multiple organ dysfunction score: Use of two strategies. Med Decis Making 1999; 19:399–410
35. Nielsen KR, Migita R, Batra M, et al. Identifying high-risk children in the emergency department. J Intensive Care Med 2016; 31:660–666
36. Zonfrillo MR, Durbin DR, Koepsell TD, et al. Prevalence of and risk factors
for poor functioning after isolated mild traumatic brain injury in children. J Neurotrauma 2014; 31:722–727
37. Edwards JD, Houtrow AJ, Vasilevskis EE, et al. Chronic conditions among children admitted to U.S. pediatric intensive care units
: Their prevalence and impact on risk for mortality and prolonged length of stay*. Crit Care Med 2012; 40:2196–2203
38. Phua J, Ngerng W, See K, et al. Characteristics and outcomes of culture-negative versus culture-positive severe sepsis
. Crit Care 2013; 17:R202
39. Gupta S, Sakhuja A, Kumar G, et al. Culture-negative severe sepsis
: Nationwide trends and outcomes. Chest 2016; 150:1251–1259
40. Menon K, McNally JD, Zimmerman JJ, et al. Primary outcome measures in Pediatric Septic Shock Trials: A systematic review. Pediatr Crit Care Med 2017; 18:e146–e154
child; intensive care units; outcome assessment; quality of life; risk factors; sepsis
Supplemental Digital Content
©2019The Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies