Confounding by indication of the safety of de-escalation in community-acquired pneumonia: A simulation study embedded in a prospective cohort

Effect of positive microbiological testing on antibiotic de-escalation and outcomes in community-acquired pneumonia: A propensity score analysis

OBJECTIVES

The usefulness of routine microbiological testing for rationalizing antibiotic use in hospitalized patients with community-acquired pneumonia (CAP) continues to be a subject of debate. We aim to determine the effect of positive microbiological testing on antimicrobial de-escalation and clinical outcomes in CAP.

METHODS

A retrospective analysis of a prospectively collected cohort of non-immunosuppressed adults hospitalized with CAP was performed. The primary study outcome was antimicrobial de-escalation. Secondary outcomes included 30-day case-fatality rate, adverse events, and CAP recurrence. Adjustment for confounders, was performed by inverse probability weighting propensity score (IPW-PS), logistic regression and cause-specific Cox model.

RESULTS

Of 3677 patients with CAP, 1924 (52.3%) had any positive microbiological test. Antimicrobial de-escalation was performed in 648/1924 (33.7%) of patients with positive microbiological testing and in 179/1753 (10.2%) of those with non positive results. When propensity score was entered into the multivariate analysis, positive microbiological testing (Adjusted Odds Ratio [AOR] 2.59 (1.96 - 3.41) and clinical stability at day 3 (AOR 1.87; 1.45 - 2.10) were two of the main factors independently associated with antimicrobial de-escalation. After applying an adjusted cause-specific Cox model, antimicrobial de-escalation was not associated with a higher 30-day case-fatality rate (Adjusted Hazard Ratio [AHR] 0.44; 0.14 - 1.43), higher frequency of adverse events (AHR 0.77; 0.53 - 1.12) or CAP recurrence (AHR 0.77; 0.45 - 1.28).

CONCLUSIONS

Antimicrobial de-escalation was more often performed in hospitalized patients with CAP who had positive microbiological tests than in those with non positive results, and it did not adversely affect relevant clinical outcomes.

Impact of Guideline Adherence on Outcomes in Patients Hospitalized with Community-Acquired Pneumonia (CAP) in Hungary: A Retrospective Observational Study

Community-acquired pneumonia (CAP) is a leading cause of morbidity and mortality worldwide. This retrospective observational study evaluated the antibiotic prescription patterns and associations between guideline adherence and outcomes in patients hospitalized with CAP in Hungary. Main outcome measures were adherence to national and international CAP guidelines (agent choice, dose) when using empirical antibiotics, antibiotic exposure, and clinical outcomes. Demographic and clinical characteristics of patients with CAP in the 30-day mortality and 30-day survival groups were compared. Fisher’s exact test and t-test were applied to compare categorical and continuous variables, respectively. Adherence to the national CAP guideline for initial empirical therapies was 30.61% (45/147) for agent choice and 88.89% (40/45) for dose. Average duration of antibiotic therapy for CAP was 7.13 ± 4.37 (mean ± SD) days, while average antibiotic consumption was 11.41 ± 8.59 DDD/patient (range 1−44.5). Adherence to national guideline led to a slightly lower 30-day mortality rate than guideline non-adherence (15.56% vs. 16.67%, p > 0.05). In patients aged ≥ 85 years, 30-day mortality was 3 times higher than in those aged 65−84 years (30.43% vs. 11.11%). A significant difference was found between 30-day non-survivors and 30-day survivors regarding the average CRP values on admission (177.28 ± 118.94 vs. 112.88 ± 93.47 mg/L, respectively, p = 0.006) and CCI score (5.71 ± 1.85 and 4.67 ± 1.83, p = 0.012). We found poor adherence to the national and international CAP guidelines in terms of agent choice. In addition, high CRP values on admission were markedly associated with higher mortality in CAP.

[Management of Adult Community-Acquired Pneumonia and Prevention - Update 2021 - Guideline of the German Respiratory Society (DGP), the Paul-Ehrlich-Society for Chemotherapy (PEG), the German Society for Infectious Diseases (DGI), the German Society of Medical Intensive Care and Emergency Medicine (DGIIN), the German Viological Society (DGV), the Competence Network CAPNETZ, the German College of General Practitioneers and Family Physicians (DEGAM), the German Society for Geriatric Medicine (DGG), the German Palliative Society (DGP), the Austrian Society of Pneumology Society (ÖGP), the Austrian Society for Infectious and Tropical Diseases (ÖGIT), the Swiss Respiratory Society (SGP) and the Swiss Society for Infectious Diseases Society (SSI)]

The present guideline provides a new and updated concept of the management of adult patients with community-acquired pneumonia. It replaces the previous guideline dating from 2016.The guideline was worked out and agreed on following the standards of methodology of a S3-guideline. This includes a systematic literature search and grading, a structured discussion of recommendations supported by the literature as well as the declaration and assessment of potential conflicts of interests.The guideline has a focus on specific clinical circumstances, an update on severity assessment, and includes recommendations for an individualized selection of antimicrobial treatment.The recommendations aim at the same time at a structured assessment of risk for adverse outcome as well as an early determination of treatment goals in order to reduce mortality in patients with curative treatment goal and to provide palliation for patients with treatment restrictions.

The impact of a stewardship program on antibiotic administration in community-acquired pneumonia: Results from an observational before-after study

BACKGROUND

A majority of patients with community-acquired pneumonia (CAP) receive antibiotics. According to the evidence, 5-7 days of treatment should be sufficient for most patients. Many, however, are treated longer than recommended. We have previously conducted a quality improvement study to ensure guideline-conform treatment for CAP. However, the impact of the interventions on antibiotic use has not been investigated.

OBJECTIVE

To estimate the impact of an eight-month stewardship program on antibiotic use.

METHODS

We conducted a before-after study comparing a four-month baseline period with data from a corresponding follow-up period. We performed univariable and multivariable logistic regression to compare odds for ≤7 days of total antibiotic treatment, ≤3 days of intravenous treatment and the proportion of correct empiric antibiotics. As sensitivity analysis, we repeated the univariable logistic regression on a propensity score-matched cohort by using the same variables we used for adjustments in the multivariable analysis. We also performed subgroup analyses for patients stable ≤72 h of admission.

RESULTS

In total, 771 patients were included. Compared to preintervention, the unadjusted odds ratio (OR) for ≤7 days of total antibiotic treatment were 1.84 (95% CI 1.34-2.54) for the whole population and 2.08 (1.41-3.10) for the stable patients. The OR for ≤3 days of intravenous antibiotics were 1.16 (0.87-1.54) and 1.38 (0.87-2.22), respectively. The OR for correct empiric antibiotics were 1.96 (1.45-2.68) and 1.82 (1.23-2.69). Comparable results regarding all outcomes were derived from the other analyses.

CONCLUSION

The program resulted in a significantly lower overall antibiotic exposure and a higher proportion of patients treated with the recommended antibiotics without a the reduction of exposure to intravenous antibiotics significantly.

Antimicrobial de-escalation in the critically ill patient and assessment of clinical cure: the DIANA study

Purpose

The DIANA study aimed to evaluate how often antimicrobial de-escalation (ADE) of empirical treatment is performed in the intensive care unit (ICU) and to estimate the effect of ADE on clinical cure on day 7 following treatment initiation.

Methods

Adult ICU patients receiving empirical antimicrobial therapy for bacterial infection were studied in a prospective observational study from October 2016 until May 2018. ADE was defined as (1) discontinuation of an antimicrobial in case of empirical combination therapy or (2) replacement of an antimicrobial with the intention to narrow the antimicrobial spectrum, within the first 3 days of therapy. Inverse probability (IP) weighting was used to account for time-varying confounding when estimating the effect of ADE on clinical cure.

Results

Overall, 1495 patients from 152 ICUs in 28 countries were studied. Combination therapy was prescribed in 50%, and carbapenems were prescribed in 26% of patients. Empirical therapy underwent ADE, no change and change other than ADE within the first 3 days in 16%, 63% and 22%, respectively. Unadjusted mortality at day 28 was 15.8% in the ADE cohort and 19.4% in patients with no change [p = 0.27; RR 0.83 (95% CI 0.60–1.14)]. The IP-weighted relative risk estimate for clinical cure comparing ADE with no-ADE patients (no change or change other than ADE) was 1.37 (95% CI 1.14–1.64).

Conclusion

ADE was infrequently applied in critically ill-infected patients. The observational effect estimate on clinical cure suggested no deleterious impact of ADE compared to no-ADE. However, residual confounding is likely.

Electronic supplementary material

The online version of this article (10.1007/s00134-020-06111-5) contains supplementary material, which is available to authorized users.

Impact of antimicrobial de-escalation on mortality: a literature review of study methodology and recommendations for observational studies

Introduction: The safety of de-escalation of empirical antimicrobial therapy is largely based on observational data, with many reporting protective effects on mortality. As there is no plausible biological explanation for this phenomenon, it is most probably caused by confounding by indication.Areas covered: We evaluate the methodology used in observational studies on the effects of de-escalation of antimicrobial therapy on mortality. We extended the search for a recent systematic review and identified 52 observational studies. The heterogeneity in study populations was large. Only 19 (36.5%) studies adjusted for confounders and four (8%) adjusted for clinical stability during admission, all as a fixed variable. All studies had methodological limitations, most importantly the lack of adjustment for clinical stability, causing bias toward a protective effect.Expert opinion: The methodology used in studies evaluating the effects of de-escalation on mortality requires improvement. We depicted all potential confounders in a directed acyclic graph to illustrate all associations between exposure (de-escalation) and outcome (mortality). Clinical stability is an important confounder in this association and should be modeled as a time-varying variable. We recommend to include de-escalation as time-varying exposure and use inverse-probability-of-treatment weighted marginal structural models to properly adjust for time-varying confounders.