Implementing a quick Sequential (Sepsis-Related) Organ Failure Assessment sepsis screening tool: an interrupted times series study

Sensitivity and specificity of a quick sequential [Sepsis-Related] organ failure assessment sepsis screening tool

AIM

There is limited evidence on the diagnostic accuracy of a quick Sequential [Sepsis-Related] Organ Failure Assessment (qSOFA) sepsis screening (SS) tool in developing nation health settings. The aim of this study was to test the diagnostic accuracy of a qSOFA-based SS tool, and the predictive validity of the qSOFA score in hospital ward patients from Argentina.

METHODS

Prospective observational study. Patients (≥18 years, without sepsis) were recruited within 24-48 hours of admission to a 169-bed tertiary referral private hospital in Buenos Aires. The index test was the qSOFA-based SS tool, and the reference standard sepsis diagnosed at discharge blindly evaluated with reference to the Sepsis-3.

RESULTS

In 1151 patients (median age 69.9 [IQR, 29.0]); 47 (4.1%) had sepsis, 413 (35.9%) had infection and 691 (60.0%) other diagnoses at discharge. The qSOFA-based SS tool (index test) had moderate sensitivity (60%), good specificity (89%), a very low positive (19%) and very high negative (98%) predictive value for sepsis diagnosed at discharge according to the Sepsis-3 criteria (reference standard). For the same outcome, the qSOFA score in isolation had a reasonable predictive validity area under receiver operating characteristics curve 0.77 (95% CI 0.70-0.83) P < 0.001.

CONCLUSION

The qSOFA score could reasonably discriminate patients at risk of developing sepsis; qSOFA-based screening may be valuable where no screening criteria are in place.

Improving sepsis screening and care in a developing nation health setting: A description of implementation

Evidence on sepsis screening and care in developing nations is insufficient to inform implementation practices in hospital wards. The aim of this multi-method study was to describe and evaluate the implementation of a three-step intervention (sepsis screening, alert activation, care) in five wards in Argentina in 2017. The implementation involved three stages: (1) context assessment, (2) development/participation in implementation strategies, and (3) evaluation of intervention adherence. Results were variable. The context assessment (Stage 1) demonstrated the value of education, proactivity towards care and team structures. Strategies developed (Stage 2) included sepsis screening and response guide, education, team rounding, posters, champions, audit/feedback and knowledge brokering. In Stage 3, staff screened 92% patients (506/547) for sepsis at ≥60% of set times; only 33% (21/64) patients had a sepsis alert activated when needed. A similar proportion of patients who had alerts activated (n = 16, 76%) or not (n = 32, 74%) received at least one element of care. The use of implementation strategies resulted in adherence to some aspects of the intervention. Future research is needed to improve sepsis screening and alert activation and care in this setting.