Out-of-Hospital Intensive Care and Mechanical Ventilation after Sepsis

A Population-Based Study of Routine Health Insurance Data

People who have survived sepsis often suffer from long-term effects (1). Some of these patients are permanently dependent upon intensive care and mechanical ventilation. In Germany, these care services have been available in the outpatient setting since 2010 (2). The present study reports the prevalence and incidence of out-of-hospital intensive care (OOH-IC) and out-of-hospital ventilation (OOH-V) following sepsis, together with patient characteristics and outcomes up to 24 months after discharge. A point of focus is whether mortality differs between sepsis survivors without and those with OOH-IC and/or OOH-V.

Methods

The present article is a retrospective, population-based cohort study. Among insured individuals of all 11 independent regional statutory health insurance funds (AOK), patients with sepsis aged 16 and over were identified in hospital data from 2018 by applying the ICD-10-GM codes for sepsis and ICD-10-GM plus the German procedure classification (OPS) codes for organ failure (3). The first hospital stay in this period was defined as the index stay, with a 12-month pre- and a 24-month post-admission observation period. Patients with incomplete insurance status during the observation period were excluded. Occurrence and recurrence of the need for OOH-IC and OOH-V within 12 months after the index stay were analyzed for each patient. OOH-IC was identified through services recorded using the general ledger accounts “5633: Intensive care in inpatient care facilities” and “5636: Outpatient intensive care”, or through services related to intensive care, intensive nursing, ventilation, tracheostomy, or a persistent vegetative state for the duration of at least one month. The definition of OOH-V was based on the billing of medical aids from product groups 12 and 14 during months in which OOH-IC services were provided. Expenditures for therapeutic nursing interventions covered under Book XI of the German Social Code (SGB XI) and for home nursing care covered under Book V (SGB V) were aggregated based on service billing data. Mortality up to two years after discharge is illustrated using Kaplan-Meier curves with a 95% confidence interval, comparing the following parameters:

• patients without OOH-IC
• OOH-IC patients without OOH-V
• OOH-IC patients with OOH-V.

The statistical group comparison was conducted using restricted mean survival times (RMSTs), measured in days up to t* = 720 days after discharge. The RMST is the average survival time in the follow-up observation period up to time point t* (4). A two-tailed z test was used to test the null hypothesis that there was no difference in RMST between the groups.

Results

Prevalence, incidence: In 2018, 87 925 cases of sepsis and 59 355 survivors after inpatient care were identified among 26.5 million individuals insured with AOK. Twelve months after sepsis, 1002 patients (1.7%, [95% confidence interval: [1.6; 1.8]; 3.8/100 000 insured individuals) were receiving OOH-IC and 729 (1.2%, [1.1; 1.3]; 2.8/100 000 insured individuals) OOH-V. OOH-IC was most commonly provided at home (71.8% of the 1002 OOH-IC patients), followed by full inpatient long-term care facilities (21.7%), and congregate living facilities (11.8%; multiple counts possible in the event of a change of residence). Among patients without prior OOH-IC or OOH-V, the incidence of OOH-IC and OOH-V was 1.3%, [1.3; 1.3] (709/54 782) and 0.9%, [0.9; 0.9] (522/55 035), respectively.

Patient characteristics: On average, OOH-IC patients were 65.5 (standard deviation (SD) = 13.0) years old, 66.8% [63.8; 69.6] were male. Most OOH-IC patients had complex comorbidities in the year before developing sepsis, including multiple, often pulmonary, pre-existing conditions. The majority experienced respiratory sepsis (78.8%, [76.2; 81.3]) or organ failure (88.7% [86.6; 90.5]) during their index stay. Tracheostomy was performed in 55.8% [52.7; 58.8] of patients during the index stay in which the sepsis occurred.

Outcomes: During the follow-up period, patients often suffered infections (71.9% [69.0; 74.6] of OOH-IC patients) or recurrent sepsis (22.1% [19.6; 24.7]), requiring rehospitalization. Non-ventilated patients initially showed higher mortality rates, which also explains the 42.2-day shorter RMST in this group (Table). Although this is reflected in significant differences in one-year mortality, the proportion of deaths after two years was nearly identical in both groups (OOH-IC without ventilation: 54.2% [48.3; 60.0], OOH-V 53.4% [49.7; 57.0], p = 0.865), yet significantly higher than among sepsis survivors without OOH-IC (40.0 % [39.6; 40.4], Figure). In the year following sepsis, the costs of care (including total care costs covered under SGB V and care services under SGB XI) amounted to 143 876 (SD = 96 673) euros per OOH-IC patient and 162 833 (SD = 96 263) euros per patient receiving OOH-V.

Figure

Kaplan-Meier curves with 95% confidence intervals according to sepsis survivor subgroups

Table

Restricted mean survival time (RMST) in days and subgroup differences among sepsis survivors, with 95% confidence intervals (CI)

Discussion

Only a small proportion of sepsis survivors receive OOH-IC and OOH-V care. However, this group presents a high rate of re-infections and recurrent sepsis and a high long-term mortality. In 2019, around 9000 persons in Germany insured with AOK received OOH-IC (2). Our results suggest that a significant proportion of these patients had previously suffered sepsis. Considering that in Germany approximately 30% of expenditures for medical nursing care and home care were attributable to intensive care services (€1.9 billion in 2018 [2]), this type of care has substantial health economic impacts. Improved sepsis prevention, for example, through vaccination campaigns, could help prevent severe courses of pneumonia and sepsis, in particular, and so potentially also reduce the number of OOH-IC cases. The present study is an initial assessment of healthcare provision involving OOH-IC and OOH-V after sepsis. Future research should investigate individual treatment courses, weaning potential, and patient-reported quality of care (5).

Norman Rose, Antje Schwinger, Thomas Ruhnke, Antje Freytag, Sören Matzk, Mathias W. Pletz, Carolin Fleischmann-Struzek

Funding

The study was funded by the Innovation Fund of the Joint Federal Committee in Germany (grant number 01VSF21031).

Conflict of interest statement

The authors declare that no conflict of interest exists.

Manuscript received on May 6, 2025, revised version accepted on October 31, 2025

Translated from the original German by Dr. Grahame Larkin

Cite this as:
Rose N, Schwinger A, Ruhnke T, Freytag A, Matzk S, Pletz MW, Fleischmann-Struzek C: Out-of-hospital intensive care and mechanical ventilation after sepsis: A population-based study of routine health insurance data. Dtsch Arztebl Int 2026; 123: 78–9. DOI: 10.3238/arztebl.m2025.0205

References

1. Fleischmann-Struzek C, Born S, Kesselmeier M, et al.: Functional ­dependence following intensive care unit-treated sepsis: Three-year ­follow-up results from the prospective Mid-German Sepsis Cohort (MSC). Lancet Reg Health Eur 2024; 46: 101066

2. Räker M, Matzk S, Büscher A, et al.: Außerklinische Intensivpflege nach dem IPReG – eine Standortbestimmung anhand von AOK-Abrechnungsdaten. In: Jacobs K, Kuhlmey, A., Greß, S., Klauber, J., Schwinger, A., editor. Pflege-Report 2022. Berlin, Heidelberg: Springer; 2022

3. Fleischmann-Struzek C, Rose N, Ditscheid B, et al.: Understanding ­health care pathways of patients with sepsis: Protocol of a mixed-­methods analysis of health care utilization, experiences, and needs of patients with and after sepsis. BMC Health Serv Res 2024; 24: 40

4. Uno H, Claggett B, Tian L, et al.: Moving beyond the hazard ratio in quantifying the between-group difference in survival analysis. J Clin ­Oncol 2014; 32: 2380–5

5. Bachmann M, Schucher B: Nach der Intensivtherapie: Außerklinische Intensivpflege. Dtsch Med Wochenschr 2024; 149: 216–22