Long-term quality of life in necrotizing soft-tissue infection survivors: a monocentric prospective cohort study

Background Compared to other life-threatening infection survivors, long-term health-related quality of life (QOL) of patients surviving necrotizing soft-tissue infections (NSTI) and its determinants are little known. In this monocentric prospective cohort including NSTI survivors admitted between 2014 and 2017, QOL was assessed during a phone interview using the 36-Item Short-Form Health Survey (SF-36), the Hospital Anxiety and Depression (HAD), the activity of daily living (ADL), instrumental ADL (IADL) scales and the Impact of Event Scale-Revised (IES-R). The primary outcome measure was the SF-36 physical component summary (PCS). NSTI patients were compared according to intensive care unit (ICU) admission status. ICU survivors were matched on SAPS II with non-NSTI related septic shock survivors. Results Forty-nine NSTI survivors were phone-interviewed and included in the study. Median PCS was decreased compared to the reference population [− 0.97 (− 2.27; − 0.08) SD]. Previous cardiac disease was the only variable associated with PCS alteration [multivariate regression coefficient: − 8.86 (− 17.64; − 0.07), p  =  0.048]. Of NSTI survivors, 15.2% had a HAD-D score  ≥  5 and 61.2% an IES-R score  ≥  33. ICU admission was not associated with lower PCS [35.21 (25.49–46.54) versus (vs) 41.82 (24.12–51.01), p  =  0.516], but with higher IES-R score [14 (7.5–34) vs 7 (3–18), p  =  0.035] and a higher proportion of HAD-D score  ≥  5 (28.6 vs 4.0%, p  =  0.036). Compared to non-NSTI septic shock-matched controls, NSTI patients had similar PCS [33.81 (24.58; − 44.39) vs 44.87 (26.71; − 56.01), p  =  0.706] but higher HAD-D [3.5 (1–7) vs 3 (1.5–6), p  =  0.048] and IES-R scores [18 (8–35) vs 8 (3–19), p  =  0.049]. Conclusions Long-term QOL in NSTI survivors is severely impaired, similarly to that of non-NSTI septic shock patients for physical compartments, but with more frequent depressive and/or post-traumatic stress disorders. Only ICU admission and previous cardiac disease were predictive of QOL impairment. Supplementary Information The online version contains supplementary material available at 10.1186/s13613-021-00891-9.

Long-term quality-of-life (QOL) of NSTI survivors has been shown to be altered when compared to the global population [7][8][9], as has also been reported among ICU survivors overall [10][11][12]. Nevertheless, long-term QOL impairment of NSTI survivors is characterized by specific features such as chronic pain, sexual dysfunction in case of perineal involvement [13], or psychological distress due to changes in body appearance [8]. Comparing NSTI survivors to other ICU patients admitted for equally severe infections might thus allow for identifying specific needs during their follow-up. Moreover, admission characteristics, including the topographical characteristics of the infection, have not been associated with long-term QOL impairment [7,13].
The aim of this study was to evaluate long-term quality of life in necrotizing soft-tissue infection survivors, compare it to an equally severe population of other infection survivors, and finally assess its association with admission characteristics.

Aim, patients and study design
The primary objective of the current study was to evaluate long-term QOL in NSTI survivors using validated psychological instruments [14][15][16][17]. Secondary objectives were to assess the impact of need for ICU admission and of other admission characteristics on QOL. In order to evaluate the severity of QOL impairment with a more comparable population than the global population, we conducted an exposed-unexposed cohort study matching NSTI patient to non-NSTI septic shock patients.
We conducted a prospective monocenter study including all consecutive adult patients who survived a hospital admission in our tertiary care center for NSTI between January 1st 2014 and September 29th 2017 [6]. NSTI survivors were categorized as "admitted to the ICU" in case of ICU admission for NSTI at any time during their hospital stay. Patients admitted to the ICU were subsequently matched 1:1 or 1:2 on SAPS II (± 5 points) to a cohort of non-NSTI septic shock survivors from a previous study assessing QOL [18].
Patients received information during their hospital stay that data abstracted from their medical charts could be used for research purposes. Data were anonymized and compiled according to the requirements of the Commission Nationale Informatique et Liberté (Registration Number 2003722) and the study was approved by the Comité de Protection des Personnes Ile-de-France V (Reference #16165). The study has been reported according to the STROBE guidelines regarding observational cohort studies.

Phone interviews
Patients were contacted by an introductory letter explaining the nature of the study. Phone interviews were conducted from November 2017 to March 2018 by two dedicated nurses using an electronic standardized form.

Data collection
The following data were collected: age, sex, main comorbidities, SAPS II, portal of entry of sepsis in patients with septic shock unrelated to NSTI, need for invasive mechanical ventilation, and hospital length of stay. To assess patient-reported outcomes, the following questionnaires were administered during phone interviews: (i) the validated French version of the Medical Outcome Study 36-Item Short-Form Health Survey (SF-36) questionnaire; (ii) the Hospital Anxiety and Depression (HAD) scale; (iii) the Impact of Event Scale-Revised (IES-R); (iv) the level of independence evaluated with the activities of daily living (ADL) and the instrumental activities of daily living (IADL) scales; (v) a self-assessed global quality of life evaluated with a quantitative score ranging from 0 (worst value) to 100 (best value); and (vi) a selection of relevant individual and contextual determinants of health-related QOL in this population.
Topographical characteristics of NSTIs were assessed by one investigator (TU) by retrospectively analyzing photographs of affected areas. These were available for 42 of the 49 patients, with post-operative photographs also available and used in priority for 30 patients. The following NSTI features were recovered: NSTI location (limb vs trunk and/or abdomino-perineal), articular skin involvement (yes or no) and circumferential skin necrosis/debridement (yes or no) for limb NSTIs. The percentage of body surface affected was also determined using the Wallace rule of nine adjusted by estimating the size of one palm to be 1% of total body surface [19].

Outcome measures
The primary outcome measure was the Physical Component Summary (PCS) from the SF-36 questionnaire. The SF-36 is a validated, generic health-related quality of life questionnaire [20][21][22]. It comprises 36 items assessing eight health-related QOL dimensions, covering physical functioning, physical role, bodily pain, general health, vitality, social functioning, emotional role, and mental health. Physical and mental component summaries (PCS and MCS, respectively) constitute aggregates of the eight individual dimensions to provide summary scores from physically and mentally oriented subscales using weights updated for French population. The higher the PCS or MCS, the better the QOL.

Statistical analysis
Quantitative data are presented as median [Interquartile range (IQR)] or mean ± standard deviations (SD), categorical data as number (percentages). Comparisons were made using the Student's t test or the Mann-Whitney test for continuous variables, and the Chi-squared or Fisher exact tests for categorical variables, according to sample size.
QOL of NSTI survivors as a whole was first described using the crude values of QOL scores. SF-36 subdimensions were expressed as raw values (normalized on a 0-100 scale) and age-sex standardized values, using reference values of the French population [23], and expressed as SD (z-scores). SF-36 aggregate components (i.e., PCS and MCS) were computed as recommended [24] and also expressed either on a normalized scale centered on 50 representing the population norm (i.e., values > or < 50 reflecting values higher or lower than age-sex standardized French values) or as age-sex standardized values (z-scores). The association of admission characteristics with PCS and MCS was assessed by linear regression. The final model was determined by entering all variables associated with the outcome at the p < 0.10 level in univariate analysis. No backward elimination of the variables was performed. Results for continuous variables, assessed in univariate and multivariate analyses, were expressed per unit of the feature and regression coefficients [95% Confidence Interval (CI)] presented.
Exploratory analyses included comparisons between NSTI survivors admitted to the ICU and non-NSTI septic shock survivors matched on SAPS II, and comparisons between NSTI survivors who had or had not been admitted to the ICU. Analyses in the matched population were adjusted for age, sex and time elapsed between hospital discharge and phone interview, using mixed models.
Significance was defined as two-sided p value < 0.05.

Characteristics and long-term QOL of the whole study population
From January 2014 to September 2017, 144 patients were admitted to our center for NSTI. Among these, 122 (84.7%) survived hospital discharge, of whom 30 were lost to follow-up, leaving 92 patients to be contacted. Of these, 43 did not answer the phone calls, leaving 49 patients who answered the QOL questionnaire after a median time period of 1.5 [1.1-2.4] years after hospital discharge, and were included in the study (Fig. 1).
Demographics, comorbidities, clinical features and QOL assessment of the study population are presented in Table 1. Patients were predominantly males with a median age of 60 [53-70] years, and the most frequent comorbidities were diabetes (36.7%), obesity (30.6%), cardiac disease (24.5%) and immunodeficiency (24.5%). Twenty-four patients (49.0%) were admitted to the ICU, of whom 11 required mechanical ventilation and 18 developed septic shock during the course of hospital stay. Regarding NSTI characteristics, limbs were the most commonly involved area (83.7%), while the trunk and/ or the abdomino-perineal region were less frequently affected (18.4%). Median body surface affected was 4.0% [2.0-4.5] with 15.6% of circumferential lesions and 46.9% of articular involvement. Nine patients (18.4%) required limb amputations. QOL in NSTI survivors was severely impaired for physical components when compared to the global population, as assessed by the SF-36 with an agesex standardized PCS aggregate score lower than that of the reference population [36.4 (24.4-48.6), z-score = − 0.97 (− 2.27; − 0.08) SD]. Regarding mental health, and although the MCS aggregate score of NSTI survivors was even slightly higher than that of the reference population [55.6 (48.8-62.9), z-score = + 0.41 SD], 37.0% of patients had a HAD-D score ≥ 5 and 20.4% an IES-R score ≥ 33, pointing to a significant incidence of depressive and/or post-traumatic stress disorders in NSTI survivors.

Association of admission characteristics with long-term QOL
An exploratory analysis found that after multivariate adjustment of admission characteristics, only previous cardiac disease remained associated with long-term physical QOL, as assessed by the PCS of the SF-36 [multivariate regression coefficient − 8.86 (− 17.64; − 0.07), p = 0.048;  Table S1).

Comparison between NSTI patients admitted to the ICU and matched non-NSTI septic shock survivors
The 24 NSTI patients admitted to the ICU were matched according to SAPS II on a 1:1 ratio for 10 patients and a 1:2 ratio for 11 patients, to a total of 32 non-NSTI septic shock patients (see Additional file 2:  Fig. 2A). Nevertheless, after standardization for age and sex, several subdimensions including role physical (p = 0.016), vitality (p = 0.033), social functioning (p = 0.036), and mental health index (p = 0.018) were significantly lower in NSTI patients when compared to their non-NSTI counterparts (Fig. 2B).

Discussion
The main findings of this study are as follows: (1) longterm QOL in NSTI survivors was severely impaired when compared to that of the reference population, particularly for physical compartments; (2) no admission characteristic aside from previous cardiac disease, including topographical characteristics of the infection, were predictive of QOL impairment; (3) long-term QOL in NSTI survivors was relatively similar to that of non-NSTI septic shock survivors, expect for a trend towards more depressive and/or post-traumatic stress disorders; and (4) ICU admission was associated with a significant alteration in mental QOL in NSTI survivors. Compared to the reference population, the QOL of NSTI survivors was severely impaired, particularly for physical compartments. Although a relatively preserved mental health was also noted in other works using the MCS of the SF-36 [25], we found, using a more extensive array of tools such as the HAD and IES scales, that   mental health was more profoundly altered in NSTI survivors compared to non-NSTI septic shock-matched controls. Similarly, studies focusing on patient experience have revealed a severe psychosocial impact among NSTI survivors [26,27]. We previously reported similar severe alteration of long-term QOL among both purpura fulminans and epidermal necrolysis survivors [18,28]. The physical alteration among these patients was less severe than that of NSTI survivors when compared to the reference population (PCS SD, respectively, − 0.63 and − 0.44 vs − 1.59). There were no differences regarding mental health between purpura fulminans or epidermal necrolysis survivors and other septic shock patients, while for NSTI patients both HAD-D and IES-R scores were significantly higher, suggesting a more severe psychological distress in these patients. This could be due to the younger age of purpura fulminans and epidermal necrolysis patients [respectively, 43 (25-61) and 48 (35-59) years old in our previous works], as compared with Multivariate analysis for association with PCS included all variables with a p value < 0.1 by univariate linear regression analysis. One patient who completed the questionnaire did not fully answer items needed for PCS calculation and only 48 patients were thus included   Fig. 2 SF-36 questionnaire results in exposed (ICU-admitted NSTI survivors) and non-exposed (non-NSTI septic shock survivors) patients. A Comparison of the physical (PCS) and mental (PCS) component summaries of the SF-36 questionnaire between patients surviving necrotizing soft tissue infections ("ICU-admitted NSTI", dark blue) and patients surviving a septic shock from other cause ("non-NSTI septic shock", light blue). Results are centered on the value of 50 (dashed line; values > or < 50 reflected values higher or lower than age-sex standardized French values). *p values come from a mixed logistic regression with adjustment for sex, SAPS II and the time between ICU discharge and phone interview, with pair as a random effect. B Comparison of each subdomain of the SF-36 questionnaire between patients surviving necrotizing soft tissue infections ("ICU-admitted NSTI", dark blue) and patients surviving a septic shock from other cause ("non-NSTI septic shock", light blue). The "physical functioning", "role physical", "bodily pain", and "general health" subdimensions pertain to the PCS aggregate component, the "vitality", "social functioning", "mental health", and "role emotional" subdimensions pertain to the MCS aggregate component. *Stands for statistical significance with p value < 0.05 from comparing between groups the standard deviations of each score standardized for age and sex, by mixed logistic regression adjusted for sex, age and the time between ICU discharge and phone interview, with pair as a random effect. , who also have more comorbidities. This is in line with studies showing that older individuals, even among NSTI survivors, are more prone to alterations of their physical QOL, although they do not corroborate the finding that psychosocial adjustment could improve with age [29]. As for NSTI patients, ICU admission was associated with a more severely impaired QOL in epidermal necrolysis patients [28]. There was a trend for longer hospital stays in ICU-admitted compared to non-ICU admitted NSTI patients. Whether this longer, although non-significant, hospital stay or ICU admission in itself accounted for mental QOL impairment is unsettled [10]. Taken together, these results suggest that life-threatening conditions affecting the skin could be associated with a more severe long-term QOL impairment than other causes of septic shock, particularly in more frail and more severe patients. We found no association between admission characteristics and QOL impairment aside from previous cardiac disease, including topographical characteristics of the infection. Although infections of the limbs have been associated with functional limitation [7], the latter has not been associated with impaired QOL, nor has articular involvement [30]. Moreover, specific fields of QOL could be affected for perineal infections such as sexual function [13]. Results regarding the association between the percentage of body surface area affected and QOL are conflicting [7,8,30]. Age, although only associated with PCS in univariate analysis in our study, likely because of its interdependence with the variable cardiac disease, has been shown to be the admission characteristic most consistently associated with long-term QOL in other studies [8,9,[29][30][31].
The main limitations of the study are its monocentric and retrospective nature, with a relatively small sample size. Although the first may limit generalizability of our results, both admission characteristics and QOL measurement are in line with the literature [8,9,25,30]. We cannot exclude that a lack of statistical power might have accounted for the absence of significant correlation between admission characteristics and long-term QOL, as well as of significant differences in QOL between NSTI and non-NSTI ICU survivors. Not all patients were admitted to the ICU, and few had perineal NSTI, which can be associated with specific alterations of QOL [13]. This warrants further larger scale multicenter studies. All ICU admitted NSTI patients were not in shock, but were matched on SAPS II to non-NSTI-related septic shock survivors to obtain more comparable populations values > or < 50 reflected values higher or lower than age-sex standardized French values).*p values come from comparing between groups the standard deviations of each score standardized for age and sex. B Comparison of each subdomain of the SF-36 between patients surviving necrotizing soft tissue infections admitted ("ICU-admitted NSTI", black) or non-admitted to the ICU ("non-ICU-admitted NSTI", gray). The "Physical functioning", "role physical", "bodily pain", and "general health" subdimensions pertain to the PCS aggregate component, the "vitality", "social functioning", "mental health", and "role emotional" subdimensions pertain to the MCS aggregate component. *Stands for statistical significance with p value < 0.05 from comparing between groups the standard deviations of each score standardized for age and sex. MCS mental component summary; PCS physical component summary; SF-36 36-Item Short-Form Health Survey; ICU intensive care unit in terms of severity, although residual confounders may remain, such as other unrecorded comorbidities. Responders and non-responders were comparable on sex (59 vs 51% male, p = 0.440) and severity of illness as assessed by admission SAPS II [25 (18-36) vs 27 (20-34), p = 0.441], but there was a non-significant trend for non-responders to be older [58 (52-69) vs 66 (53-78) years, p = 0.074], and a selection bias is possible. Finally, we did not assess microbiological characteristics or therapeutic interventions and their association with QOL, although in a small-scale study E. coli infections and an important number of surgical excisions were predictors of poor outcome [30]. Similarly, marital status has been associated with QOL for NSTI survivors, an association we did not evaluate [8,29].
Our results are supportive of a more intense longterm follow-up of NSTI patients including physiotherapy and psychological support to detect and manage both physical and mental alterations of QOL, particularly in patients whose condition was severe upon admission. These findings finally highlight the need for further research in the little explored field of NSTI, using patientcentered outcomes.

Conclusion
Long-term QOL in NSTI survivors was severely impaired when compared to the reference population, particularly for physical compartments. It was relatively similar to that of non-NSTI septic shock patients, expect for a trend towards more frequent depressive and/or post-traumatic stress disorders. Apart from ICU admission and previous cardiac disease, no admission characteristic, including topographical characteristics of the infection, were predictive of QOL impairment.