 |
Medication Safety Officers Society 4100 Members Strong A society of healthcare professionals dedicated to improving medication safety in healthcare organizations |
Prusaczyk B, Burke RE. It’s time for the field of geriatrics to invest in implementation science. BMJ Qual Saf 2023;32:700-703.
In this article, the affiliation ‘Central and North West London NHS foundation Trust’ has been added to Simon Conroy and the funding statement has been updated to acknowledge the funding from the NHS Elect and Central and North West London NHS foundation Trust.
doi: 10.1136/bmjqs-2023-016263corr1
Tackling overuse in healthcare is now more necessary than ever. Movements such as Choosing Wisely and Preventing Overdiagnosis have highlighted that some healthcare services offer no added value and may even cause harm to patients. Estimates of overdiagnosis and overtreatment vary widely between services, providers and regions.1 2 Overuse is a persistent challenge in high-income countries and is increasingly recognised in low-income settings.3 Action is needed to prevent patient harm, reduce resource waste and preserve the limited time of healthcare professionals. In addition, since healthcare services have a significant environmental impact, minimising overuse can contribute to achieving climate goals.
De-implementation scienceTo accelerate the reduction of overuse, robust de-implementation science is essential.4 This field studies the drivers, strategies and processes involved in reducing or eliminating ineffective, unnecessary or harmful healthcare practices, and in replacing them with evidence-based, high-value alternatives. Rigorous...
Sepsis remains a leading cause of morbidity and mortality worldwide.1 The use of artificial intelligence (AI), and particularly machine-learning (ML) approaches, to predict which patients are at risk for sepsis in the hospital may improve patient-centred outcomes through early recognition and timely antibiotics. Yet, despite major interest in the use of ML applications in sepsis care, there are only a handful of successful examples of model implementation that save lives through early detection.2 3 The high cost and extensive system architecture required to test and implement novel ML applications have limited many institutions’ abilities to bring these models to the bedside. Unfortunately, this has resulted in a preponderance of studies on model development rather than implementation and reliance on proprietary models disseminated to health systems without validation or testing. One well-known sepsis predictive model developed by an electronic health record vendor (Epic Systems,...
Low-value care (LVC) describes practices that persist in healthcare, despite being ineffective, inefficient or causing harm. Several determinants for the provision of LVC have been identified, but understanding how these factors influence professionals’ decisions, individually and jointly, is a necessary next step to guide deimplementation.
MethodsA factorial survey experiment was employed using vignettes that presented hypothetical medical scenarios among 593 Swedish primary care physicians. Each vignette varied systematically by factors such as patient age, patient request for the LVC, physician’s perception of this practice, practice cost to the primary care centre and time taken to deliver it. For each scenario, we measured the reported likelihood of providing the LVC. We collected information on the physician’s worry about missing a serious illness.
ResultsPatient requests and physicians’ positive perceptions of the practice were the factors that increased the reported likelihood of providing LVC the most (by 14 and 13 percentage points (pp), respectively). When the LVC was low in cost or not time-consuming, patient requests further boosted the likelihood of provision by 29 and 18 pp. In contrast, credible evidence against the LVC reduced the role of patient requests by 11 pp. Physicians’ fear of missing a serious illness was linked with higher reported probability of providing LVC, and the credibility of the evidence against the LVC reduced the role of this concern.
ConclusionsThe findings highlight that patient requests enhance the role of many determinants, while the credibility of evidence diminishes the impact of others. Overall, these findings point to the relevance of increased clinician knowledge about LVC, tools for patient communication and the use of decision support tools to reduce the uncertainty in decision-making.
To identify bias in using a single machine learning (ML) sepsis prediction model across multiple hospitals and care locations; evaluate the impact of six different bias mitigation strategies and propose a generic modelling approach for developing best-performing models.
MethodsWe developed a baseline ML model to predict sepsis using retrospective data on patients in emergency departments (EDs) and wards across nine hospitals. We set model sensitivity at 70% and determined the number of alerts required to be evaluated (number needed to evaluate (NNE), 95% CI) for each case of true sepsis and the number of hours between the first alert and timestamped outcomes meeting sepsis-3 reference criteria (HTS3). Six bias mitigation models were compared with the baseline model for impact on NNE and HTS3.
ResultsAcross 969 292 admissions, mean NNE for the baseline model was significantly lower for EDs (6.1 patients, 95% CI 6 to 6.2) than for wards (7.5 patients, 95% CI 7.4 to 7.5). Across all sites, median HTS3 was 20 hours (20–21) for wards vs 5 (5–5) for EDs. Bias mitigation models significantly impacted NNE but not HTS3. Compared with the baseline model, the best-performing models for NNE with reduced interhospital variance were those trained separately on data from ED patients or from ward patients across all sites. These models generated the lowest NNE results for all care locations in seven of nine hospitals.
ConclusionsImplementing a single sepsis prediction model across all sites and care locations within multihospital systems may be unacceptable given large variances in NNE across multiple sites. Bias mitigation methods can identify models demonstrating improved performance across most sites in reducing alert burden but with no impact on the length of the prediction window.
There is limited evidence from antimicrobial stewardship programmes in less-resourced settings. This study aimed to improve the quality of antibacterial prescriptions by mitigating overuse and promoting the use of narrow-spectrum agents in intensive care units (ICUs) in a middle-income country.
MethodsWe established a quality improvement collaborative (QIC) model involving nine Argentine ICUs over 11 months with a 16-week baseline period (BP) and a 32-week implementation period (IP). Our intervention package included audits and feedback on antibacterial use, facility-specific treatment guidelines, antibacterial timeouts, pharmacy-based interventions and education. The intervention was delivered in two learning sessions with three action periods along with coaching support and basic quality improvement training.
ResultsWe included 912 patients, 357 in BP and 555 in IP. The latter had higher APACHE II (17 (95% CI: 12 to 21) vs 15 (95% CI: 11 to 20), p=0.036), SOFA scores (6 (95% CI: 4 to 9) vs 5 (95% CI: 3 to 8), p=0.006), renal failure (41.6% vs 33.1%, p=0.009), sepsis (36.1% vs 31.6%, p<0.001) and septic shock (40.0% vs 33.8%, p<0.001). The days of antibacterial therapy (DOT) were similar between the groups (change in the slope from BP to IP 28.1 (95% CI: –17.4 to 73.5), p=0.2405). There were no differences in the antibacterial defined daily dose (DDD) between the groups (change in the slope from BP to IP 43.9, (95% CI: –12.3 to 100.0), p=0.1413).
The rate of antibacterial de-escalation based on microbiological culture was higher during the IP (62.0% vs 45.3%, p<0.001).
The infection prevention control (IPC) assessment framework was increased in eight ICUs.
ConclusionImplementing an antimicrobial stewardship program in ICUs in a middle-income country via a QIC demonstrated success in improving antibacterial de-escalation based on microbiological culture results, but not on DOT or DDD. In addition, eight out of nine ICUs improved their IPC Assessment Framework Score.
Early mobilisation of intensive care unit (ICU) patients has been recommended in clinical practice guidelines. Therefore, the Japanese universal health insurance system introduced an additional fee for early mobilisation and/or rehabilitation, which can be claimed by hospitals when starting rehabilitation of ICU patients within 48 hours after their ICU admission. However, the effect of this fee is unknown.
ObjectiveTo measure the proportion of ICU patients who received early rehabilitation and the impact on length of ICU stay, the length of hospital stay and discharged to home after the introduction of the financial incentive (additional fee for early mobilisation and/or rehabilitation).
Design/methodsWe included patients who were admitted to ICU within 2 days of hospitalisation between April 2016 and January 2020. We conducted interrupted time series analyses to assess the effects of the introduction of the financial incentive.
ResultsThe proportion of patients who received early rehabilitation immediately increased after the introduction of the financial incentive (rate ratio (RR) 1.293, 95% CI 1.240 to 1.349). The RR for proportion of patients received early rehabilitation was 1.008 (95% CI 1.005 to 1.011) in the period after the introduction of the financial incentive compared with period before its introduction. There was no statistically significant change in the mean length of ICU stay, the mean length of hospital stay and the proportion of patients who were discharged to home.
ConclusionAfter the introduction of the financial incentive, the proportion of ICU patients who received early rehabilitation increased. However, the effects of the financial incentive on the length of ICU stay, the length of hospital stay and the proportion of patients who were discharged to home were limited.
Creating and sustaining an institutional climate conducive to patient and health worker safety is a critical element of successful multimodal hand hygiene improvement strategies aimed at achieving best practices. Repeated WHO global surveys indicate that the institutional safety climate consistently ranks the lowest among various interventions.
MethodsTo develop an international expert consensus on research agenda priorities related to the role of institutional safety climate within the context of a multimodal hand hygiene improvement strategy, we conducted a structured consensus process involving a purposive sample of international experts. A preliminary list of research priorities was formulated following evidence mapping, and subsequently refined through a modified Delphi consensus process involving two rounds. In round 1, survey respondents were asked to rate the importance of each research priority. In round 2, experts reviewed round 1 ratings to reach a consensus (defined as ≥70% agreement) on the final prioritised items to be included in the research agenda. The research priorities were then reviewed and finalised by members of the WHO Technical Advisory Group on Hand Hygiene Research in Healthcare.
ResultsOf the 57 invited participants, 50 completed Delphi round 1 (88%), and 48 completed round 2 (96%). Thirty-six research priority statements were included in round 1 across five thematic categories: (1) safety climate; (2) personal accountability for hand hygiene; (3) leadership; (4) patient participation and empowerment and (5) religion and traditions. In round 1, 75% of the items achieved consensus, with 9 statements carried forward to round 2, leading to a final set of 31 prioritised research statements.
ConclusionThis research agenda can be used by researchers, clinicians, policy-makers and funding bodies to address gaps in hand hygiene improvement within the context of an institutional safety climate, thereby enhancing patient and health worker safety globally.
The way that data are presented can influence quality and safety initiatives. Time-series charts highlight changes but do not clarify whether data lie outside expected variation. Statistical process control (SPC) charts make this distinction and have been demonstrated to be effective in supporting hospital initiatives. To improve the uptake of the SPC methodology by hospitals in England, a training intervention was created. The current study evaluates the effectiveness of that training against the background of a wider national initiative to encourage the adoption of SPC charts.
MethodsA parallel cluster randomised trial was conducted with 16 English NHS hospitals. Half were randomised to the training intervention and half to the control. The primary analysis compares the difference in use of SPC charts within hospital board papers in a postrandomisation period (adjusting for baseline use). Trainees completed feedback forms with Likert scale and open-ended items.
ResultsFifteen hospitals participated across the study arms. SPC chart use increased in both intervention and control hospitals between the baseline and postrandomisation period (29 and 30 percentage points, respectively). There was no statistically significant difference between the intervention and control hospitals in use of SPC charts in the postrandomisation period (average absolute difference 9% (95% CI –34% to 52%). In the feedback forms, 93.9% (n=31/33) of trainees affirmed learning and 97.0% (n=32/33) had formed an intention to change their behaviour.
ConclusionsControl chart use increased in both intervention and control hospitals. This is consistent with a rising tide and/or contamination effect, such that the culture of control chart use is spreading across hospitals in England. Further research is needed to support hospitals implementing SPC training initiatives and to link SPC implementation to quality and safety outcomes. Such research could support future quality and safety initiatives nationally and internationally.
Trial registration numberThe stepped-wedge cluster randomised trial is a popular design in implementation and health services research. All clusters, such as clinics or hospitals, start in the control state, and gradually switch over to treatment in a random order until all clusters have received the intervention. The design allows for the incorporation of an experiment into the gradual roll-out of an intervention across clusters. However, the traditional stepped-wedge layout may not be the best choice in many scenarios. In this article, we discuss modifications to the stepped-wedge design that maintain a staggered roll-out, but which may improve some key characteristics. We consider improving the timing of implementation periods, reducing the volume of data collection and allowing for the recruitment of clusters over the course of the trial.
