RESUMEN
BACKGROUND: We previously implemented the SQuID protocol (subcutaneous insulin in diabetic ketoacidosis [DKA]) demonstrating safe, effective treatment of low- to moderate-severity DKA in a non-intensive care unit setting. Since success and sustainability of interventions rely on staff buy-in, we assessed acceptability of SQuID among emergency department (ED) and inpatient clinicians. METHODS: We conducted a cross-sectional study in an urban academic hospital (March 2023-November 2023), surveying ED nurses (RNs) and physicians (MDs) and floor RNs and MDs treating patients on SQuID via emailed survey links. Clinicians could only take the survey once. We used Sekhon's Theoretical Framework of Acceptability, validated for staff acceptability of a new intervention, assessing eight domains with 5-point Likert responses. Clinicians were asked about prior experience with SQuID, and we assessed ED MD and RN preference (SQuID vs. intravenous [IV] insulin). Surveys included free-text boxes for comments. We present descriptive statistics including proportions with 95% confidence interval and medians with interquartile ranges (IQRs) and conducted thematic analysis of free-text comments. RESULTS: Our overall response rate (107/133) was 80% (34/42 ED RNs, 13/16 floor RNs, 47/57 ED MDs, 13/17 floor MDs), with first-time users of SQuID ranging from 7.7% (hospitalist MDs) to 35.3% (ED RNs) of participants. ED clinicians preferred SQuID over IV insulin (67% vs. 12%, 21% no preference). Acceptability was high across all domains and clinician types (median 4, IQR 4-5). Overall percentage of positive responses (4s and 5s) across domains was 92% (ED RNs [89%], floor RNs [89%], ED MDs [97%], floor MDs [87%]). We identified several themes among participant comments. CONCLUSIONS: Acceptability was high across clinician types; 65% of ED clinicians preferred SQuID to IV insulin. Clinicians liked SQuID (affective attitude), found it easy to use (burden), were confident in its use (self-efficacy), felt that it improved outcomes (perceived effectiveness), found that it was fair to patients (ethicality), found that it made sense (intervention coherence), and found that it did not interfere with other activities (opportunity cost).
RESUMEN
OBJECTIVE: We previously demonstrated safe treatment of low- to moderate-severity (LTM) diabetic ketoacidosis (DKA) using the SQuID protocol (subcutaneous insulin in DKA) in a non-intensive care unit (ICU) observation setting, with decreased emergency department length of stay (EDLOS). Here, we expand eligibility to include sicker patients and admission to a regular medical floor and collected more detailed clinical data in a near-real-time fashion. METHODS: This is a real-world, prospective, observational cohort study in an urban academic hospital (March 4, 2023-March 4, 2024). LTM DKA patients were treated with IV insulin (floor or ICU) or on SQuID. We compare fidelity (time to glargine and dextrose-containing fluids), safety (rescue dextrose for hypoglycemia), effectiveness (time to anion gap closure, time on protocol), and operational efficiency (time to bed request, EDLOS, and ICU admission rate since implementation of the protocol). RESULTS: Of 84 patients with LTM DKA, 62 (74%) of were treated with SQuID and 22 (26%) with IV insulin. Fidelity was high in both groups. Rescue dextrose was required in five (8%) versus four (18%) patients, respectively (difference 9%, -31% to 10%). Compared to the IV insulin group, time to anion gap was 1.4 h shorter (95% CI -3.4 to 0.2 h) and time on protocol was 10.4 h shorter (95% CI -22.3 to -5.0 h) in SQuID patients. Median EDLOS was lower in the SQuID cohort 9.8 h (IQR 6.0-13.6) than the IV floor cohort 18.3 h (IQR 13.4-22.0 h), but longer than the overall IV insulin cohort. Since inception of SQuID, ICU admission rate in LTM DKA has decreased from 54% to under 21%. CONCLUSIONS: In this single-center study, we observed excellent fidelity, equivalent or superior safety, and clinical and operational effectiveness with SQuID compared to IV insulin. The SQuID protocol has become the de facto default pathway for treatment of LTM DKA. Since inception of SQuID, ICU admissions in LTM DKA have decreased 33%.
RESUMEN
BACKGROUND: Studies using fast-acting subcutaneous (SQ) insulin analogs in diabetic ketoacidosis (DKA) have demonstrated efficacy, safety, and cost-effectiveness, allowing treatment of mild-to-moderate (MTM)-severity DKA patients in non-intensive care unit (ICU) settings. However, emergency department (ED)-based studies are few, with limited exploration of impacts on operational metrics. METHODS: We implemented the SQuID (Subcutaneous Insulin in Diabetic Ketoacidosis) protocol for adults with MTM-severity DKA in an urban academic ED, collecting data from August 1, 2021, to February 28, 2022. We examined fidelity (frequency of required q2h glucose checks), safety (proportion of patients administered rescue dextrose for hypoglycemia), and ED length of stay (EDLOS) for the SQuID cohort compared to patients (non-ICU) treated with a traditional insulin infusion. We also examined ICU admission rate among MTM-severity DKA patients after introduction of SQuID to two historical control periods (pre-intervention and pre-COVID). We used Mann-Whitney U to test for differences in EDLOS distributions, bootstrapped (n = 1000) confidence intervals (CIs) for EDLOS median differences, and the two-sample z-test for differences in ICU admissions. RESULTS: We identified 177 MTM-severity DKA patients in the study period (78 SQuID, 99 traditional cohort) and 163 preintervention and 161 pre-COVID historical control patients. Fidelity to the SQuID pathway was good, with glucose checks exceeding the q2-h requirement. We found no difference in the proportion of rescue dextrose administration compared to the traditional pathway. We observed significant reductions in median EDLOS for the SQuID cohort compared to the traditional cohort during the study period (-3.0, 95% CI -8.5 to -1.4), the preintervention period (-1.4, 95% CI -3.1 to -0.1), and the pre-COVID control period (-3.6, 95% CI -7.5 to -1.8). CONCLUSIONS: In this single-center study at an academic ED, treatment of patients with MTM-severity DKA with a SQ insulin protocol was effective, demonstrated equivalent safety, and reduced ED length of stay.
