1913. Clinical Performance of the qSOFA Score Among Non-ICU Inpatients With Infection at a Tertiary Hospital in Jamaica From 2015 to 2016

Abstract

BackgroundSepsis is common and catastrophic. The usefulness of the qSOFA score has been questioned. Thus far, data on the validity of the instrument have been derived from developed countries. The generalizability to developing countries is unknown. This study aimed to ascertain how “qSOFA” predicted death and need for intensive care in patients at a tertiary hospital in Jamaica.MethodsSeven hundred fifty-two patients admitted between January 2015 and December 2016 with a physician determined diagnosis of infection were randomly selected from the electronic medical database. The details of the first 48 hours of their admission were reviewed.ResultsMost patients were middle-aged females who remained in hospital for an average of 9 days and were managed by Internal Medicine. Two of the most common sites of infection (respiratory and gastrointestinal) were also the two sites associated with the highest risk of death or requiring intensive care. 126 (17%) had a qSOFA score ≥2 at presentation, 4 (0.5%) persons died, and 32 (4%) required admission to ICU. Many more patients met the SIRS criteria than qSOFA at presentation (66% vs. 17%). Meeting the SIRS criteria, however, was not significantly associated with death or needing intensive care. On the other hand, those with a positive qSOFA at presentation were three times more likely to die or need intensive care (OR 3.03; 95% CI 1.1,8.9; P = 0.04). The qSOFA score detected these patients, with a high degree of specificity (84%), especially when utilized at presentation (OR 3.03; 95% CI 1.03–8.92; P = 0.04) and 48 hours after (OR 2.24, 95% CI 0.94–5.37, P = 0.07). The sensitivity of the qSOFA score was poor (39%), but this was improved to 100% when combined with the SIRS score at presentation. There was a suggestion that this combined score also offered the best prognostic accuracy with an AUROC of 0.74 (95% CI 0.66–0.81) when compared with the qSOFA score (AUROC –0.68, 95% CI 0.60–0.76) or SIRS criteria alone (AUROC –0.71, 95% CI 0.63–0.79). However, there was significant overlap of the curves and the differences were not significant. ConclusionAmong non-ICU inpatients with infection, the qSOFA score is useful for predicting death and the need for ICU. However, its utility is improved when used alongside and not instead of the SIRS criteria.Disclosures All authors: No reported disclosures.