Modelling interrupted time series to evaluate prevention and control of infection in healthcare

Does health voucher intervention increase antenatal consultations and skilled birth attendances in Cameroon? Results from an interrupted time series analysis

BACKGROUND

Limited access to health services during the antenatal period and during childbirth, due to financial barriers, is an obstacle to reducing maternal and child mortality. To improve the use of health services in the three regions of Cameroon, which have the worst reproductive, maternal, neonatal, child and adolescent health indicators, a health voucher project aiming to reduce financial barriers has been progressively implemented since 2015 in these three regions. Our research aimed to assess the impact of the voucher scheme on first antenatal consultation (ANC) and skilled birth attendance (SBA).

METHODS

Routine aggregated data by month over the period January 2013 to May 2018 for each of the 33 and 37 health facilities included in the study sample were used to measure the effect of the voucher project on the first ANC and SBA, respectively. We estimated changes attributable to the intervention in terms of the levels of outcome indicators immediately after the start of the project and over time using an interrupted time series regression. A meta-analysis was used to obtain the overall estimates.

RESULTS

Overall, the voucher project contributed to an immediate and statistically significant increase, one month after the start of the project, in the monthly number of ANCs (by 26%) and the monthly number of SBAs (by 57%). Compared to the period before the start of the project, a statistically significant monthly increase was observed during the project implementation for SBAs but not for the first ANCs. The results at the level of health facilities (HFs) were mixed. Some HFs experienced an improvement, while others were faced with the status quo or a decrease.

CONCLUSIONS

Unlike SBAs, the voucher project in Cameroon had mixed results in improving first ANCs. These limited effects were likely the consequence of poor design and implementation challenges.

Abortion Provision and Delays to Care in a Clinic Network in Washington State After Dobbs

Importance

The Supreme Court decision Dobbs v Jackson Women's Health Organization (Dobbs) overturned federal protections to abortion care and altered the reproductive health care landscape. Thus far, aggregated state-level data reveal increases in the number of abortions in states where abortion is still legal, but there is limited information on delays to care and changes in the characteristics of people accessing abortion in these states after Dobbs.

Objective

To examine changes in abortion provision and delays to care after Dobbs.

Design, Setting, and Participants

Retrospective cohort study of all abortions performed at an independent, high-volume reproductive health care clinic network in Washington state from January 1, 2017, to July 31, 2023. Using an interrupted time series, the study assessed changes in abortion care after Dobbs.

Exposure

Abortion care obtained after (June 24, 2022, to July 31, 2023) vs before (January 1, 2017, to June 23, 2022) Dobbs.

Main Outcome and Measure

Primary outcomes included weekly number of abortions and out-of-state patients and weekly average of gestational duration (days) and time to appointment (days).

Results

Among the 18 379 abortions during the study period, most were procedural (13 192 abortions [72%]) and funded by public insurance (11 412 abortions [62%]). The mean (SD) age of individuals receiving abortion care was 28.5 (6.44) years. Following Dobbs, the number of procedural abortions per week increased by 6.35 (95% CI, 2.83-9.86), but then trended back toward pre-Dobbs levels. The number of out-of-state patients per week increased by 2 (95% CI, 1.1-3.6) and trends remained stable. The average gestational duration per week increased by 6.9 (95% CI, 3.6-10.2) days following Dobbs, primarily due to increased gestations of procedural abortions. The average gestational duration among out-of-state patients did not change following Dobbs, but it did increase by 6 days for in-state patients (5.9; 95% CI, 3.2-8.6 days). There were no significant changes in time to appointment.

Conclusions and Relevance

These findings provide a detailed picture of changes in abortion provision and delays to care after Dobbs in a state bordering a total ban state. In this study, more people traveled from out of state to receive care and in-state patients sought care a week later in gestation. These findings can inform interventions and policies to improve access for all seeking abortion care.

The effect of psychiatric decision unit services on inpatient admissions and mental health presentations in emergency departments: an interrupted time series analysis from two cities and one rural area in England

AIMS

High-quality evidence is lacking for the impact on healthcare utilisation of short-stay alternatives to psychiatric inpatient services for people experiencing acute and/or complex mental health crises (known in England as psychiatric decision units [PDUs]). We assessed the extent to which changes in psychiatric hospital and emergency department (ED) activity were explained by implementation of PDUs in England using a quasi-experimental approach.

METHODS

We conducted an interrupted time series (ITS) analysis of weekly aggregated data pre- and post-PDU implementation in one rural and two urban sites using segmented regression, adjusting for temporal and seasonal trends. Primary outcomes were changes in the number of voluntary inpatient admissions to (acute) adult psychiatric wards and number of ED adult mental health-related attendances in the 24 months post-PDU implementation compared to that in the 24 months pre-PDU implementation.

RESULTS

The two PDUs (one urban and one rural) with longer (average) stays and high staff-to-patient ratios observed post-PDU decreases in the pattern of weekly voluntary psychiatric admissions relative to pre-PDU trend (Rural: -0.45%/week, 95% confidence interval [CI] = -0.78%, -0.12%; Urban: -0.49%/week, 95% CI = -0.73%, -0.25%); PDU implementation in each was associated with an estimated 35-38% reduction in total voluntary admissions in the post-PDU period. The (urban) PDU with the highest throughput, lowest staff-to-patient ratio and shortest average stay observed a 20% (-20.4%, CI = -29.7%, -10.0%) level reduction in mental health-related ED attendances post-PDU, although there was little impact on long-term trend. Pooled analyses across sites indicated a significant reduction in the number of voluntary admissions following PDU implementation (-16.6%, 95% CI = -23.9%, -8.5%) but no significant (long-term) trend change (-0.20%/week, 95% CI = -0.74%, 0.34%) and no short- (-2.8%, 95% CI = -19.3%, 17.0%) or long-term (0.08%/week, 95% CI = -0.13, 0.28%) effects on mental health-related ED attendances. Findings were largely unchanged in secondary (ITS) analyses that considered the introduction of other service initiatives in the study period.

CONCLUSIONS

The introduction of PDUs was associated with an immediate reduction of voluntary psychiatric inpatient admissions. The extent to which PDUs change long-term trends of voluntary psychiatric admissions or impact on psychiatric presentations at ED may be linked to their configuration. PDUs with a large capacity, short length of stay and low staff-to-patient ratio can positively impact ED mental health presentations, while PDUs with longer length of stay and higher staff-to-patient ratios have potential to reduce voluntary psychiatric admissions over an extended period. Taken as a whole, our analyses suggest that when establishing a PDU, consideration of the primary crisis-care need that underlies the creation of the unit is key.

Interpretation of coefficients in segmented regression for interrupted time series analyses

Background Segmented regression, a common model for interrupted time series (ITS) analysis, primarily utilizes two equation parametrizations. Interpretations of coefficients vary between the two segmented regression parametrizations, leading to occasional user misinterpretations. Methods To illustrate differences in coefficient interpretation between two common parametrizations of segmented regression in ITS analysis, we derived analytical results and present an illustration evaluating the impact of a smoking regulation policy in Italy using a publicly accessible dataset. Estimated coefficients and their standard errors were obtained using two commonly used parametrizations for segmented regression with continuous outcomes. We clarified coefficient interpretations and intervention effect calculations. Results Our investigation revealed that both parametrizations represent the same model. However, due to differences in parametrization, the immediate effect of the intervention is estimated differently under the two approaches. The key difference lies in the interpretation of the coefficient related to the binary indicator for intervention implementation, impacting the calculation of the immediate effect. Conclusions Two common parametrizations of segmented regression represent the same model but have different interpretations of a key coefficient. Researchers employing either parametrization should exercise caution when interpreting coefficients and calculating intervention effects.

An Evaluation of the Impact of an OPEN Stewardship Generated Feedback Intervention on Antibiotic Prescribing among Primary Care Veterinarians in Canada and Israel

An interrupted time-series study design was implemented to evaluate the impact of antibiotic stewardship interventions on antibiotic prescribing among veterinarians. A total of 41 veterinarians were enrolled in Canada and Israel and their prescribing data between 2019 and 2021 were obtained. As an intervention, veterinarians periodically received three feedback reports comprising feedback on the participants' antibiotic prescribing and prescribing guidelines. A change in the level and trend of antibiotic prescribing after the administration of the intervention was compared using a multi-level generalized linear mixed-effect negative-binomial model. After the receipt of the first (incidence rate ratios [IRR] = 0.88; 95% confidence interval (CI): 0.79, 0.98), and second (IRR = 0.85; 95% CI: 0.75, 0.97) feedback reports, there was a reduced prescribing rate of total antibiotic when other parameters were held constant. This decline was more pronounced among Israeli veterinarians compared to Canadian veterinarians. When other parameters were held constant, the prescribing of critical antibiotics by Canadian veterinarians decreased by a factor of 0.39 compared to that of Israeli veterinarians. Evidently, antibiotic stewardship interventions can improve antibiotic prescribing in a veterinary setting. The strategy to sustain the effect of feedback reports and the determinants of differences between the two cohorts should be further explored.

Comparison of statistical methods used to meta-analyse results from interrupted time series studies: an empirical study

BACKGROUND

The Interrupted Time Series (ITS) is a robust design for evaluating public health and policy interventions or exposures when randomisation may be infeasible. Several statistical methods are available for the analysis and meta-analysis of ITS studies. We sought to empirically compare available methods when applied to real-world ITS data.

METHODS

We sourced ITS data from published meta-analyses to create an online data repository. Each dataset was re-analysed using two ITS estimation methods. The level- and slope-change effect estimates (and standard errors) were calculated and combined using fixed-effect and four random-effects meta-analysis methods. We examined differences in meta-analytic level- and slope-change estimates, their 95% confidence intervals, p-values, and estimates of heterogeneity across the statistical methods.

RESULTS

Of 40 eligible meta-analyses, data from 17 meta-analyses including 282 ITS studies were obtained (predominantly investigating the effects of public health interruptions (88%)) and analysed. We found that on average, the meta-analytic effect estimates, their standard errors and between-study variances were not sensitive to meta-analysis method choice, irrespective of the ITS analysis method. However, across ITS analysis methods, for any given meta-analysis, there could be small to moderate differences in meta-analytic effect estimates, and important differences in the meta-analytic standard errors. Furthermore, the confidence interval widths and p-values for the meta-analytic effect estimates varied depending on the choice of confidence interval method and ITS analysis method.

CONCLUSIONS

Our empirical study showed that meta-analysis effect estimates, their standard errors, confidence interval widths and p-values can be affected by statistical method choice. These differences may importantly impact interpretations and conclusions of a meta-analysis and suggest that the statistical methods are not interchangeable in practice.

The impact of mass vaccination policy and control measures on lumpy skin disease cases in Thailand: insights from a Bayesian structural time series analysis

Introduction

In 2021, Thailand reported the highest incidence of lumpy skin disease (LSD) outbreaks in Asia. In response to the widespread outbreaks in cattle herds, the government's livestock authorities initiated comprehensive intervention measures, encompassing control strategies and a national vaccination program. Yet, the efficacy of these interventions remained unevaluated. This research sought to assess the nationwide intervention's impact on the incidence of new LSD cases through causal impact analysis.

Methods

Data on weekly new LSD cases in Thailand from March to September 2021 was analyzed. The Bayesian structural time series (BSTS) analysis was employed to evaluate the causal relationship between new LSD cases in the pre-intervention phase (prior to the vaccination campaign) and the post-intervention phase (following the vaccination campaign). The assessment involved two distinct scenarios, each determined by the estimated effective intervention dates. In both scenarios, a consistent decline in new LSD cases was observed after the mass vaccination initiative, while other control measures such as the restriction of animal movement, insect control, and the enhancement of the active surveillance approach remained operational throughout the pre-intervention and the post-intervention phases.

Results and discussion

According to the relative effect results obtained from scenario A and B, it was observed that the incidence of LSD cases exhibited reductions of 119% (95% Credible interval [CrI]: -121%, -38%) and 78% (95% CrI: -126, -41%), respectively. The BSTS results underscored the significant influence of these interventions, with a Bayesian one-sided tail-area probability of p < 0.05. This model-based study provides insight into the application of BSTS in evaluating the impact of nationwide LSD vaccination based on the national-level data. The present study is groundbreaking in two respects: it is the first study to quantify the causal effects of a mass vaccination intervention on the LSD outbreak in Thailand, and it stands as the only endeavor of its kind in the Asian context. The insights collected from this study hold potential value for policymakers in Thailand and other countries at risk of LSD outbreaks.

Coronavirus disease 2019 (COVID-19) vaccination rates and staffing shortages among healthcare personnel in nursing homes before, during, and after implementation of mandates for COVID-19 vaccination among 15 US jurisdictions, National Healthcare Safety Network, June 2021-January 2022

OBJECTIVE

To examine temporal changes in coverage with a complete primary series of coronavirus disease 2019 (COVID-19) vaccination and staffing shortages among healthcare personnel (HCP) working in nursing homes in the United States before, during, and after the implementation of jurisdiction-based COVID-19 vaccination mandates for HCP.

SAMPLE AND SETTING

HCP in nursing homes from 15 US jurisdictions.

DESIGN

We analyzed weekly COVID-19 vaccination data reported to the Centers for Disease Control and Prevention's National Healthcare Safety Network from June 7, 2021, through January 2, 2022. We assessed 3 periods (preintervention, intervention, and postintervention) based on the announcement of vaccination mandates for HCP in 15 jurisdictions. We used interrupted time-series models to estimate the weekly percentage change in vaccination with complete primary series and the odds of reporting a staffing shortage for each period.

RESULTS

Complete primary series vaccination among HCP increased from 66.7% at baseline to 94.3% at the end of the study period and increased at the fastest rate during the intervention period for 12 of 15 jurisdictions. The odds of reporting a staffing shortage were lowest after the intervention.

CONCLUSIONS

These findings demonstrate that COVID-19 vaccination mandates may be an effective strategy for improving HCP vaccination coverage in nursing homes without exacerbating staffing shortages. These data suggest that mandates can be considered to improve COVID-19 coverage among HCP in nursing homes to protect both HCP and vulnerable nursing home residents.

Evaluation of statistical methods used to meta-analyse results from interrupted time series studies: A simulation study

Interrupted time series (ITS) are often meta-analysed to inform public health and policy decisions but examination of the statistical methods for ITS analysis and meta-analysis in this context is limited. We simulated meta-analyses of ITS studies with continuous outcome data, analysed the studies using segmented linear regression with two estimation methods [ordinary least squares (OLS) and restricted maximum likelihood (REML)], and meta-analysed the immediate level- and slope-change effect estimates using fixed-effect and (multiple) random-effects meta-analysis methods. Simulation design parameters included varying series length; magnitude of lag-1 autocorrelation; magnitude of level- and slope-changes; number of included studies; and, effect size heterogeneity. All meta-analysis methods yielded unbiased estimates of the interruption effects. All random effects meta-analysis methods yielded coverage close to the nominal level, irrespective of the ITS analysis method used and other design parameters. However, heterogeneity was frequently overestimated in scenarios where the ITS study standard errors were underestimated, which occurred for short series or when the ITS analysis method did not appropriately account for autocorrelation. The performance of meta-analysis methods depends on the design and analysis of the included ITS studies. Although all random effects methods performed well in terms of coverage, irrespective of the ITS analysis method, we recommend the use of effect estimates calculated from ITS methods that adjust for autocorrelation when possible. Doing so will likely to lead to more accurate estimates of the heterogeneity variance.

Nationwide implementation of personalized outcomes forecasts to support physical therapists in treating patients with intermittent claudication: Protocol for an interrupted time series study

INTRODUCTION

Shared decision-making is the cornerstone of patient-centered care. However, evidence suggests that the application of shared decision-making in physical therapy practice is limited. To elicit shared decision-making and thereby potentially improve patient outcomes for patients with intermittent claudication, we developed a decision support system. This decision support system provides personalized outcomes forecasts that visualize the estimated walking distance of an individual patient. We hypothesize that personalized outcomes forecasts can support physical therapists in personalizing care to the needs and priorities of the individual patient to improve therapy outcomes.

RESEARCH OBJECTIVES

The primary aim is to evaluate the impact of personalized outcomes forecasts for patients with intermittent claudication to optimize personalized treatment. Second, this study aims to evaluate the process of implementation.

METHODS

This study uses a prospective interrupted time series (ITS) design. Participating physical therapists are divided into four clusters. Every month of the study period, a new cluster will be invited to begin using the decision support system. We aim to include data of 11,250 newly referred patients for physical therapy treatment. All therapists associated with a network of specialized therapists (Chronic CareNet) and patients treated by these therapists are eligible to participate. The decision support system, called the KomPas, makes use of personalized outcomes forecasts, which visualize the estimated outcome of supervised exercise therapy for an individual patient with intermittent claudication. Personalized outcomes forecasts are developed using a neighbors-based approach that selects patients similar to the index patient (a.k.a. neighbors) from a large database. Outcomes to evaluate impact of implementation are patients' functional and maximal walking distance, quality of life and shared decision-making. Process evaluation will be measured in terms of utilization efficacy, including the outcomes dropout rate and reasons to (not) use the personalized outcomes forecasts. Data will be routinely collected through two online systems: the Chronic CareNet Quality system, and the website logs of the decision support system. Additionally, observations and semi-structured interviews will be conducted with a small subset of therapists.

ETHICS

Formal medical ethical approval by the Medical Research Ethics Committees United 'MEC-U' was not required for this study under Dutch law (reference number 2020-6250).

Effect estimates can be accurately calculated with data digitally extracted from interrupted time series graphs

Interrupted time series (ITS) studies are frequently used to examine the impact of population-level interventions or exposures. Systematic reviews with meta-analyses including ITS designs may inform public health and policy decision-making. Re-analysis of ITS may be required for inclusion in meta-analysis. While publications of ITS rarely provide raw data for re-analysis, graphs are often included, from which time series data can be digitally extracted. However, the accuracy of effect estimates calculated from data digitally extracted from ITS graphs is currently unknown. Forty-three ITS with available datasets and time series graphs were included. Time series data from each graph was extracted by four researchers using digital data extraction software. Data extraction errors were analysed. Segmented linear regression models were fitted to the extracted and provided datasets, from which estimates of immediate level and slope change (and associated statistics) were calculated and compared across the datasets. Although there were some data extraction errors of time points, primarily due to complications in the original graphs, they did not translate into important differences in estimates of interruption effects (and associated statistics). Using digital data extraction to obtain data from ITS graphs should be considered in reviews including ITS. Including these studies in meta-analyses, even with slight inaccuracy, is likely to outweigh the loss of information from non-inclusion.

Intervening on women's health for rural young breast cancer survivors: A study protocol

INTRODUCTION

From diagnosis to post-treatment, many young breast cancer survivors (YBCS) experience infertility, limited contraception choices, concern about pregnancy safety, and menopausal symptoms. Clinical guidelines recommend oncofertility care (counseling and/or clinical services that meet fertility, contraception, pregnancy health and/or menopausal symptom management needs) throughout the cancer care continuum. However, significant oncofertility care gaps exist in rural, community oncology settings.

MATERIALS AND METHODS

We describe the design of an interrupted time series, effectiveness-implementation hybrid clinical trial that evaluates a multi-component intervention to improve YBCS engagement in oncofertility care. The intervention is comprised of 1) oncology clinic-based oncofertility needs screen; 2) a women's health survivorship care plan in Spanish and English; 3) remote patient navigation; and 4) telehealth oncofertility consultation. During the pre-intervention period (12 months), usual care will be delivered. During the intervention period (15 months), the multi-component intervention will be implemented at two rural oncology clinics with largely Latina, Spanish-speaking populations. The primary outcome of YBCS (n = 135) engagement in oncofertility care will be collected from medical record review. We will also collect validated patient-reported outcomes. Informed by the Exploration Preparation Implementation Sustainment (EPIS) implementation science framework, we will integrate qualitative and quantitative data to explore whether and how the intervention was effective, acceptable, appropriate, and delivered with fidelity.

DISCUSSION

Our overall goal is to speed implementation of a scalable oncofertility care intervention for YBCS in underserved areas to reduce disparities and improve reproductive health and quality of life.

TRIAL REGISTRATION

Clinicaltrials.gov Identifier: NCT05414812.

The Impact of the Swedish Care Coordination Act on Hospital Readmission and Length-of-Stay among Multi-Morbid Elderly Patients: A Controlled Interrupted Time Series Analysis

Coordinating follow-up care after discharge from hospital is critical to ensuring good outcomes for patients, but is difficult when multiple care providers are involved. In 2018, Sweden adopted the Care Coordination Act, which modified economic incentives to reduce discharge delays and mandated a discharge planning process for patients requiring post-discharge social- or primary care services. This study evaluates the impact of this reform on hospital length-of-stay and unplanned readmissions among multi-morbid elderly patients. Interrupted time series analysis of all in-patient care episodes involving multi-morbid elderly patients in Sweden from 2015 - 2019 (n = 2 386 039) was performed. Secondary analyses using case-mix adjustment and controlled interrupted time series analysis were employed to assess for bias. Average length of stay decreased during the post-reform period, corresponding to 248 521 saved care days. Unplanned readmissions meanwhile increased, corresponding to 7 572 excess unplanned readmissions. While reductions in length-of-stay were concentrated among patients targeted by the reform, increases in readmission rates were similar in patients not targeted by the reform, indicating potential confounding. The reform thus appears to have achieved its goal of decreasing in-patient length of stay, but a robust effect on readmissions, outpatient visits, or mortality was not found. This may be due to lackluster implementation or an ineffective mandated intervention.

Transformational Performance Improvement: Why Is Progress so Slow?

In this chapter, we identify three distinct transformational performance improvement (TPI) approaches commonly used to redesign work processes in health care organizations. We describe the unique components or tools that each approach uses to improve the delivery of health services. We also summarize what is empirically known about the effectiveness of each TPI approach according to systematic reviews and recent studies published in the peer-reviewed literature. Based on examination of this research, we discuss what knowledge is still needed to strengthen the evidence for whole system transformation. This involves the use of conceptual frameworks to assess and guide implementation efforts, and facilitators and barriers to change as revealed in a recent evaluation of one major initiative, the Lean Enterprise Transformation (LET) at the Veterans Health Administration. The analysis suggests ways in which TPI facilitators can be developed and barriers reduced to improve the effectiveness and sustainability of quality initiatives. Finally, we discuss appropriate study designs to evaluate TPI interventions that may strengthen the evidence for their effectiveness in real world practice settings.

Least Squares and Robust Rank-Based Double Bootstrap Analyses for Time-Series Intervention Designs

Time-series intervention designs that include two or more phases have been widely discussed in the healthcare literature for many years. A convenient model for the analysis of these designs has a linear model part (to measure changes in level and trend) plus a second part that measures the random error structure; the error structure is assumed to follow an autoregressive time-series process. Traditional generalized linear model approaches widely used to estimate this model are less than satisfactory because they tend to provide substantially biased intervention tests and confidence intervals. We describe an updated version of the original double bootstrap approach that was developed by McKnight et al. (2000) to correct for this problem. This updated analysis and a new robust version were recently implemented in an R package (McKean & Zhang, 2018). The robust method is insensitive to outliers and problems associated with common departures from normality in the error distribution. Monte Carlo studies as well as published data are used to demonstrate the properties of both versions. The R code required to perform the analyses is provided and illustrated.

Responses to COVID-19 with probabilistic programming

The COVID-19 pandemic left its unique mark on the twenty-first century as one of the most significant disasters in history, triggering governments all over the world to respond with a wide range of interventions. However, these restrictions come with a substantial price tag. It is crucial for governments to form anti-virus strategies that balance the trade-off between protecting public health and minimizing the economic cost. This work proposes a probabilistic programming method to quantify the efficiency of major initial non-pharmaceutical interventions. We present a generative simulation model that accounts for the economic and human capital cost of adopting such strategies, and provide an end-to-end pipeline to simulate the virus spread and the incurred loss of various policy combinations. By investigating the national response in 10 countries covering four continents, we found that social distancing coupled with contact tracing is the most successful policy, reducing the virus transmission rate by 96% along with a 98% reduction in economic and human capital loss. Together with experimental results, we open-sourced a framework to test the efficacy of each policy combination.

An improved method for analysis of interrupted time series (ITS) data: accounting for patient heterogeneity using weighted analysis

Interrupted time series (ITS) design is commonly used to evaluate the impact of interventions in healthcare settings. Segmented regression (SR) is the most commonly used statistical method and has been shown to be useful in practical applications involving ITS designs. Nevertheless, SR is prone to aggregation bias, which leads to imprecision and loss of power to detect clinically meaningful differences. The objective of this article is to present a weighted SR method, where variability across patients within the healthcare facility and across time points is incorporated through weights. We present the methodological framework, provide optimal weights associated with data at each time point and discuss relevant statistical inference. We conduct extensive simulations to evaluate performance of our method and provide comparative analysis with the traditional SR using established performance criteria such as bias, mean square error and statistical power. Illustrations using real data is also provided. In most simulation scenarios considered, the weighted SR method produced estimators that are uniformly more precise and relatively less biased compared to the traditional SR. The weighted approach also associated with higher statistical power in the scenarios considered. The performance difference is much larger for data with high variability across patients within healthcare facilities. The weighted method proposed here allows us to account for the heterogeneity in the patient population, leading to increased accuracy and power across all scenarios. We recommend researchers to carefully design their studies and determine their sample size by incorporating heterogeneity in the patient population.

Changes in cesarean section rate before and after the end of the Korean Value Incentive Program

BACKGROUND

The Korean government implemented a value incentive program providing incentives to providers based on C-section rates, with the rates being publicized. The program ended in 2014 after the administration decided that the effects of the incentive program were limited. In this report, we analyzed changes in C-section rates with the value incentive program.

METHODS

The analysis used claim data from Korea's National Health Insurance. The study period (2011-2016) was divided into two phases: before and after the program. This study included 95 providers that were tertiary or general hospitals having more than 200 deliveries per year during the study period. The dependent variable was the risk-adjusted C-section rate. Independent variables included time and hospital characteristics such as hospital type, district, and ownership. Interrupted time series analysis was performed to analyze the data.

RESULTS

Our results showed that risk-adjusted C-section rates increased immediately after the end of the incentive program for C-sections. The immediate effect of intervention, a change of 1.73% (P < .05), was statistically significant, as was the trend after intervention, at 0.21% (P < .0001). The slope showed an increase after the intervention to 0.25% per medical institution, which was contrary to the trend of the preintervention decline (negative slope).

CONCLUSION

Risk-adjusted C-section rates increased immediately after the discontinuation of a value incentive program. Tertiary hospitals showed greater increases in C-section rates than general hospitals after the intervention.

Changes in Migration and Mortality Among Patients With Kidney Failure in Puerto Rico After Hurricane Maria

Importance

On September 20, 2017, one of the most destructive hurricanes in US history made landfall in Puerto Rico. Anecdotal reports suggest that many persons with kidney failure left Puerto Rico after Hurricane Maria; however, empirical estimates of migration and health outcomes for this population are scarce.

Objective

To assess the changes in migration and mortality among patients with kidney failure in need of dialysis treatment in Puerto Rico after Hurricane Maria.

Design, Setting, and Participants

This cross-sectional study used an interrupted time-series design of 6-month mortality rates and migration of 11 652 patients who received hemodialysis or peritoneal dialysis care in Puerto Rico before Hurricane Maria (before October 1, 2017) and/or during and after Hurricane Maria (on/after October 1, 2017). Data analyses were performed from February 12, 2019, to June 16, 2022..

Main Outcomes and Measures

Number of unique persons dialyzed in Puerto Rico per quarter; receipt of dialysis treatment outside Puerto Rico per quarter; and 6-month mortality rate per person-quarter for all persons undergoing dialysis.

Exposures

Hurricane Maria.

Results

The entire study sample comprised 11 652 unique persons (mean [SD] age, 59 [14.7] years; 7157 [61.6%] men and 4465 [38.4%] women; 10 675 [91.9%] Hispanic individuals). There were 9022 patients with kidney failure and dialysis treatment before and 5397 patients after Hurricane Maria. Before the hurricane, the mean quarterly number of unique persons dialyzed in Puerto Rico was 2834 per quarter (95% CI, 2771-2897); afterwards it dropped to 261 (95% CI, -348 to -175; relative change, 9.2%). The percentage of persons who had 1 or more dialysis sessions outside of Puerto Rico in the next quarter following a previous dialysis in Puerto Rico was 7.1% before Hurricane Maria (95% CI, 4.8 to 9.3). There was a significant increase of 5.8 percentage points immediately after the hurricane (95% CI, 2.7 to 9.0). The 6-month mortality rate per person-quarter was 0.08 (95% CI, 0.08 to 0.09), and there was a nonsignificant increase in level of mortality rates and a nonsignificant decreasing trend in mortality rates.

Conclusions and Relevance

The findings of this cross-sectional study suggest there was a significant increase in the number of people receiving dialysis outside of Puerto Rico after Hurricane Maria. However, no significant differences in mortality rates before and after the hurricane were found, which may reflect disaster emergency preparedness among dialysis facilities and the population with kidney failure, as well as efforts from other stakeholders.

Road to Better Work-Life Balance? Lean Redesigns and Daily Work Time among Primary Care Physicians

PURPOSE

To assess the impact of Lean primary care redesigns on the amount of time that physicians spent working each day.

METHODS

This observational study was based on 92 million time-stamped Epic® EHR access logs captured among 317 primary care physicians in a large ambulatory care delivery system. Seventeen clinic facilities housing 46 primary care departments were included for study. We conducted interrupted time series analysis to monitor changes in physician work patterns over 6 years. Key measures included total daily work time; time spent on "desktop medicine" outside the exam room; time spent with patients during office visits; time still working after clinic, i.e., after seeing the last patient each day; and remote work time.

RESULTS

The amount of time that physicians spent on desktop EHR activities throughout the day, including after clinic hours, decreased by 10.9% (95% CI: -22.2, -2.03) and 8.3% (95% CI: -13.8, -2.12), respectively, during the first year of Lean implementation. Total daily work hours among physicians, which included both desktop activity and time in office visits, decreased by 20% (95% CI: -29.2, -9.60) by the third year of Lean implementation.

CONCLUSIONS

These findings suggest that Lean redesign may be associated with time savings for primary care physicians. However, since this was an observational analysis, further study is warranted (e.g., randomized trial) -to determine the impact of Lean interventions on physician work experiences.

Impact of the COVID-19 pandemic on perinatal care and outcomes in the United States: An interrupted time series analysis

BACKGROUND

Hospitals quickly adapted perinatal care to mitigate SARS-CoV-2 transmission at the onset of the COVID-19 pandemic. The objective of this study was to estimate the impact of pandemic-related hospital policy changes on perinatal care and outcomes in one region of the United States.

METHODS

This interrupted time series analysis used retrospective data from consecutive singleton births at 15 hospitals in the Pacific Northwest from 2017 to 2020. The primary outcomes were those hypothesized to be affected by pandemic-related hospital policies and included labor induction, epidural use, oxytocin augmentation, mode of delivery, and early discharge (<48 hours after cesarean and <24 hours after vaginal births). Secondary outcomes included preterm birth, severe maternal morbidity, low 5-minute Apgar score, neonatal intensive care unit (NICU) admission, and 30-day readmission. Segmented Poisson regression models estimated the outcome level shift changes after the pandemic onset, controlling for underlying trends, seasonality, and stratifying by parity.

RESULTS

No statistically significant changes were detected in intrapartum interventions or mode of delivery after onset of the pandemic. Early discharge increased for all births following cesarean and vaginal birth. Newborn readmission rates increased but only among nulliparas (aRR: 1.49, 95%CI: 1.17, 1.91). Among multiparas, decreases were observed in preterm birth (aRR: 0.90, 95%CI: 0.84, 0.96), low 5-minute Apgar score (aRR: 0.75, 95%CI: 0.68, 0.81), and term NICU admission rates (aRR: 0.85, 95%CI: 0.80, 0.91).

CONCLUSIONS

Increases in early discharge and newborn readmission rates among nulliparas suggest a need for more postpartum support during the pandemic. Decreases in preterm birth and term NICU admission among multiparas may have implications beyond the pandemic and deserve further study.

Improving health care facility birth rates in Rorya District, Tanzania: a multiple baseline trial

BACKGROUND

Rates of maternal mortality and morbidity in Africa remain unacceptably high, as many women deliver at home, without access to skilled birth attendants and life-saving medications. In rural Tanzania, women face significant barriers accessing health care facilities for their deliveries.

METHODS

From January 2017 to February 2019 we conducted a multiple baseline (interrupted time series) trial within the four divisions of Rorya District, Tanzania. We collected baseline data, then sequentially introduced a complex intervention in each of the divisions, in randomized order, over 3 month intervals. We allowed for a 6 month transition period to avoid contamination between the pre- and post-intervention periods. The intervention included using community health workers to educate about safe delivery, distribution of birth kits with misoprostol, and a transport subsidy for women living a distance from the health care facility. The primary outcome was the health facility birth rate, while the secondary outcomes were the rates of antenatal and postpartum care and postpartum hemorrhage. Outcomes were analyzed using fixed effects segmented logistic regression, adjusting for age, marital status, education, and parity. Maternal and baby morbidity/mortality were analyzed descriptively.

RESULTS

We analyzed data from 9565 pregnant women (2634 before and 6913 after the intervention was implemented). Facility births increased from 1892 (71.8%) before to 5895 (85.1%) after implementation of the intervention. After accounting for the secular trend, the intervention was associated with an immediate increase in the odds of facility births (OR = 1.51, 95% CI 1.14 to 2.01, p = 0.0045) as well as a small gradual effect (OR = 1.03 per month, 95% CI 1.00 to 1.07, p = 0.0633). For the secondary outcomes, there were no statistically significant immediate changes associated with the intervention. Rates of maternal and baby morbidity/mortality were low and similar between the pre- and post-implementation periods.

CONCLUSIONS

Access to health care facilities can be improved through implementation of education of the population by community health workers about the importance of a health care facility birth, provision of birth kits with misoprostol to women in late pregnancy, and access to a transport subsidy for delivery for women living at a distance from the health facility.

CLINICAL TRIALS REGISTRATION

NCT03024905 19/01/2017.

Methodological systematic review recommends improvements to conduct and reporting when meta-analysing interrupted time series studies

OBJECTIVES

Interrupted Time Series (ITS) are a type of non-randomised design commonly used to evaluate public health policy interventions, and the impact of exposures, at the population level. Meta-analysis may be used to combine results from ITS across studies (in the context of systematic reviews) or across sites within the same study. We aimed to examine the statistical approaches, methods, and completeness of reporting in reviews that meta-analyse results from ITS.

STUDY DESIGN AND SETTINGS

Eight electronic databases were searched to identify reviews (published 2000-2019) that meta-analysed at least two ITS. Characteristics of the included reviews, the statistical methods used to analyse the ITS and meta-analyse their results, effect measures, and risk of bias assessment tools were extracted.

RESULTS

Of the 4213 identified records, 54 reviews were included. Nearly all reviews (94%) used two-stage meta-analysis, most commonly fitting a random effects model (69%). Among the 41 reviews that re-analysed the ITS, linear regression (39%) and ARIMA (20%) were most commonly used; 38% adjusted for autocorrelation. The most common effect measure meta-analysed was an immediate level-change (46/54). Reporting of the statistical methods and ITS characteristics was often incomplete.

CONCLUSION

Improvement is needed in the conduct and reporting of reviews that meta-analyse results from ITS.

Comparison of ARIMA, ES, GRNN and ARIMA–GRNN hybrid models to forecast the second wave of COVID-19 in India and the United States

As acute infectious pneumonia, the coronavirus disease-2019 (COVID-19) has created unique challenges for each nation and region. Both India and the United States (US) have experienced a second outbreak, resulting in a severe disease burden. The study aimed to develop optimal models to predict the daily new cases, in order to help to develop public health strategies. The autoregressive integrated moving average (ARIMA) models, generalised regression neural network (GRNN) models, ARIMA–GRNN hybrid model and exponential smoothing (ES) model were used to fit the daily new cases. The performances were evaluated by minimum mean absolute per cent error (MAPE). The predictive value with ARIMA (3, 1, 3) (1, 1, 1)₁₄ model was closest to the actual value in India, while the ARIMA–GRNN presented a better performance in the US. According to the models, the number of daily new COVID-19 cases in India continued to decrease after 27 May 2021. In conclusion, the ARIMA model presented to be the best-fit model in forecasting daily COVID-19 new cases in India, and the ARIMA–GRNN hybrid model had the best prediction performance in the US. The appropriate model should be selected for different regions in predicting daily new cases. The results can shed light on understanding the trends of the outbreak and giving ideas of the epidemiological stage of these regions.

Evaluation of statistical methods used in the analysis of interrupted time series studies: a simulation study

BACKGROUND

Interrupted time series (ITS) studies are frequently used to evaluate the effects of population-level interventions or exposures. However, examination of the performance of statistical methods for this design has received relatively little attention.

METHODS

We simulated continuous data to compare the performance of a set of statistical methods under a range of scenarios which included different level and slope changes, varying lengths of series and magnitudes of lag-1 autocorrelation. We also examined the performance of the Durbin-Watson (DW) test for detecting autocorrelation.

RESULTS

All methods yielded unbiased estimates of the level and slope changes over all scenarios. The magnitude of autocorrelation was underestimated by all methods, however, restricted maximum likelihood (REML) yielded the least biased estimates. Underestimation of autocorrelation led to standard errors that were too small and coverage less than the nominal 95%. All methods performed better with longer time series, except for ordinary least squares (OLS) in the presence of autocorrelation and Newey-West for high values of autocorrelation. The DW test for the presence of autocorrelation performed poorly except for long series and large autocorrelation.

CONCLUSIONS

From the methods evaluated, OLS was the preferred method in series with fewer than 12 points, while in longer series, REML was preferred. The DW test should not be relied upon to detect autocorrelation, except when the series is long. Care is needed when interpreting results from all methods, given confidence intervals will generally be too narrow. Further research is required to develop better performing methods for ITS, especially for short series.

Implementing Lean Quality Improvement in Primary Care: Impact on Efficiency in Performing Common Clinical Tasks

BACKGROUND

Many primary care practices have adopted Lean techniques to reduce the amount of time spent completing routine tasks. Few studies have evaluated both immediate and sustained impacts of Lean to improve this aspect of primary care work efficiency.

OBJECTIVE

To examine 3-year impacts of Lean implementation on the amount of time taken for physicians to complete common clinical tasks.

DESIGN

Non-randomized stepped wedge with segmented regression and interrupted time series analysis (January 2011-December 2016).

PARTICIPANTS

A total of 317 physician-led teams in 46 primary care departments in a large ambulatory care delivery system.

INTERVENTION

Lean redesign was initiated in one pilot site followed by system-wide spread across all primary care departments. Redesigns included standardization of exam room equipment and supplies, streamlining of call management processes, care team co-location, and team management of the electronic inbox.

MEASURES

Time-stamped EHR tracking of physicians' completion time for 4 common tasks: (1) office visit documentation and closure of patient charts; (2) telephone call resolution; (3) prescription refill renewal; and (4) response to electronic patient messages.

RESULTS

After Lean implementation, we found decreases in the amount of time to complete: office visit documentation (- 29.2% [95% CI: - 44.2, - 10.1]), telephone resolution (- 22.2% [95% CI: - 38.1, - 2.27]), and renewal of prescription refills (- 2.96% per month [95% CI: - 4.21, - 1.78]). These decreases were sustained over several years. Response time to electronic patient messages did not change significantly.

CONCLUSIONS

Lean redesigns led to improvements in timely completion of 3 out of 4 common clinical tasks. Our findings support the use of Lean techniques to engage teams in routine aspects of patient care. More research is warranted to understand the mechanisms by which Lean promotes quality improvement and effectiveness of care team workflows.

Evaluation of an OPEN Stewardship generated feedback intervention to improve antibiotic prescribing among primary care veterinarians in Ontario, Canada and Israel: protocol for evaluating usability and an interrupted time-series analysis

INTRODUCTION

Antimicrobial resistance (AMR) impacts the health and well-being of animals, affects animal owners both socially and economically, and contributes to AMR at the human and environmental interface. The overuse and/or inappropriate use of antibiotics in animals has been identified as one of the most important drivers of the development of AMR in animals. Effective antibiotic stewardship interventions such as feedback can be adopted in veterinary practices to improve antibiotic prescribing. However, the provision of dedicated financial and technical resources to implement such systems are challenging. The newly developed web-based Online Platform for Expanding Antibiotic Stewardship (OPEN Stewardship) platform aims to automate the generation of feedback reports and facilitate wider adoption of antibiotic stewardship. This paper describes a protocol to evaluate the usability and usefulness of a feedback intervention among veterinarians and assess its impact on individual antibiotic prescribing.

METHODS AND ANALYSIS

Approximately 80 veterinarians from Ontario, Canada and 60 veterinarians from Israel will be voluntarily enrolled in a controlled interrupted time-series study and their monthly antibiotic prescribing data accessed. The study intervention consists of targeted feedback reports generated using the OPEN Stewardship platform. After a 3-month preintervention period, a cohort of veterinarians (treatment cohort, n=120) will receive three feedback reports over the course of 6 months while the remainder of the veterinarians (n=20) will be the control cohort. A survey will be administered among the treatment cohort after each feedback cycle to assess the usability and usefulness of various elements of the feedback report. A multilevel negative-binomial regression analysis of the preintervention and postintervention antibiotic prescribing of the treatment cohort will be performed to evaluate the impact of the intervention.

ETHICS AND DISSEMINATION

Research ethics board approval was obtained at each participating site prior to the recruitment of the veterinarians. The study findings will be disseminated through open-access scientific publications, stakeholder networks and national/international meetings.

Quasi-experimental evaluation of text-based crisis patterns in youth following Hurricane Florence in the Carolinas, 2018

IMPORTANCE

Crisis text lines have proven to be an effective and low-cost means for delivering texting-based mental health support to youth. Yet there has been limited research examining the use of these services in capturing the psychological impact on youth affected by a weather-related disaster.

OBJECTIVE

This ecologic study examined changes in help-seeking behavior for adolescents and young adults in North and South Carolina, USA, before and after Hurricane Florence (2018).

DESIGN AND MAIN OUTCOMES

A retrospective, interrupted time-series design was used to examine pre- and post-hurricane changes in crisis text volume among youth help seekers in the Carolinas for the following outcomes: (1) text for any reason; (2) stress & anxiety; (3) depression; and (4) suicidal thoughts.

RESULTS

Results showed an immediate and sustained increase in crisis texts for stress/anxiety and suicidal thoughts in the six weeks following Florence. Overall, an immediate 15% increase in crisis texts for anxiety/stress (SE = 0.05, p = .005) and a 17% increase in suicidal thoughts (SE = 0.07, p = .02) occurred during the week of the storm. Text volume for anxiety/stress increased 17% (SE = 0.08, p = .005) and 23% for suicidal ideation (SE = 0.08, p = .01) in the 6-week post-hurricane period. Finally, forecast models revealed observed text volume for all mental health outcomes was higher than expected in the 6 weeks post-Florence.

CONCLUSIONS AND RELEVANCE

A low-cost, crisis texting platform provided 24/7 mental health support available to young people in the Carolinas impacted by Hurricane Florence. These findings highlight a new application for text-based crisis support services to address the mental health consequences in youth following a weather-related disaster, as well as the potential for these types of crisis platforms to measure situational awareness in impacted communities.

Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol

Background: Systematic reviews are used to inform healthcare decision making. In reviews that aim to examine the effects of organisational, policy change or public health interventions, or exposures, evidence from interrupted time series (ITS) studies may be included. A core component of many systematic reviews is meta-analysis, which is the statistical synthesis of results across studies. There is currently a lack of guidance informing the choice of meta-analysis methods for combining results from ITS studies, and there have been no studies examining the meta-analysis methods used in practice. This study therefore aims to describe current meta-analysis methods used in a cohort of reviews of ITS studies.

Methods: We will identify the 100 most recent reviews (published between 1 January 2000 and 11 October 2019) that include meta-analyses of ITS studies from a search of eight electronic databases covering several disciplines (public health, psychology, education, economics). Study selection will be undertaken independently by two authors. Data extraction will be undertaken by one author, and for a random sample of the reviews, two authors. From eligible reviews we will extract details at the review level including discipline, type of interruption and any tools used to assess the risk of bias / methodological quality of included ITS studies; at the meta-analytic level we will extract type of outcome, effect measure(s), meta-analytic methods, and any methods used to re-analyse the individual ITS studies. Descriptive statistics will be used to summarise the data.

Conclusions: This review will describe the methods used to meta-analyse results from ITS studies. Results from this review will inform future methods research examining how different meta-analysis methods perform, and ultimately, the development of guidance.

Patient experiences after implementing lean primary care redesigns

OBJECTIVE

To examine the effect of Lean primary care redesigns on patient satisfaction with care and timeliness of care received.

DATA/SETTING

We used patient surveys and time-stamped electronic health record (EHR) data in a large ambulatory care system.

DESIGN

Lean-based changes to clinical spaces and care team workflows were implemented in one pilot site and then scaled to all primary care departments across the system. Redesigns included standardizing equipment and patient education materials in examination rooms, streamlining call management functions, co-locating physician and medical assistant dyads in a shared workspace, and creating new care team workflows. We used a non-randomized stepped-wedge study design and segmented regression with interrupted time series analysis to examine Lean impacts on patient outcomes.

DATA COLLECTION

We analyzed patient satisfaction ratings and wait times as documented by the EHR. These longitudinal data were collected for 317 physician-led teams in 46 primary care departments from January 2011 to December 2016.

PRINCIPAL FINDINGS

After implementation of Lean redesigns, patients reported a 44.8 percent increase in satisfaction with the adequacy of time spent with care providers during office visits (P < .05). They also reported 71.6 percent higher satisfaction with their care provider's ability to listen to their concerns, and a 55.4 percent increase in perceived staff helpfulness at the visit (P < .01). Based on monthly EHR data, the amount of time elapsed between a patient request for a routine appointment and the scheduled visit day decreased from baseline by an average 2 percent per month (P < .01). On the day of the visit, patient wait times to be seen also decreased gradually by an average 1.2 percent per month (P < .05).

CONCLUSIONS

Patient experiences of care after Lean implementations have not been widely studied in primary care settings. We found that Lean redesign yielded improvements that may strengthen clinical operations while enhancing value for patients.

Recurrent time-to-event models with ordinal outcomes

A model to accommodate time-to-event ordinal outcomes was proposed by Berridge and Whitehead. Very few studies have adopted this approach, despite its appeal in incorporating several ordered categories of event outcome. More recently, there has been increased interest in utilizing recurrent events to analyze practical endpoints in the study of disease history and to help quantify the changing pattern of disease over time. For example, in studies of heart failure, the analysis of a single fatal event no longer provides sufficient clinical information to manage the disease. Similarly, the grade/frequency/severity of adverse events may be more important than simply prolonged survival in studies of toxic therapies in oncology. We propose an extension of the ordinal time-to-event model to allow for multiple/recurrent events in the case of marginal models (where all subjects are at risk for each recurrence, irrespective of whether they have experienced previous recurrences) and conditional models (subjects are at risk of a recurrence only if they have experienced a previous recurrence). These models rely on marginal and conditional estimates of the instantaneous baseline hazard and provide estimates of the probabilities of an event of each severity for each recurrence over time. We outline how confidence intervals for these probabilities can be constructed and illustrate how to fit these models and provide examples of the methods, together with an interpretation of the results.

Creating effective interrupted time series graphs: Review and recommendations

Introduction

Interrupted Time Series (ITS) studies may be used to assess the impact of an interruption, such as an intervention or exposure. The data from such studies are particularly amenable to visual display and, when clearly depicted, can readily show the short‐ and long‐term impact of an interruption. Further, well‐constructed graphs allow data to be extracted using digitizing software, which can facilitate their inclusion in systematic reviews and meta‐analyses.

Aim

We provide recommendations for graphing ITS data, examine the properties of plots presented in ITS studies, and provide examples employing our recommendations.

Methods and results

Graphing recommendations from seminal data visualization resources were adapted for use with ITS studies. The adapted recommendations cover plotting of data points, trend lines, interruptions, additional lines and general graph components. We assessed whether 217 graphs from recently published (2013‐2017) ITS studies met our recommendations and found that 130 graphs (60%) had clearly distinct data points, 100 (46%) had trend lines, and 161 (74%) had a clearly defined interruption. Accurate data extraction (requiring distinct points that align with axis tick marks and labels that allow the points to be interpreted) was possible in only 72 (33%) graphs.

Conclusion

We found that many ITS graphs did not meet our recommendations and could be improved with simple changes. Our proposed recommendations aim to achieve greater standardization and improvement in the display of ITS data, and facilitate re‐use of the data in systematic reviews and meta‐analyses.

Application of data visualization recommendations can improve quality of interrupted time series graphs.

Well‐designed graphs accurately depict time series data, any impact of the interruption, and the results of the analysis.

Well‐designed graphs facilitate data extraction for use in systematic reviews and reproducibility.

An assessment of graphs included in interrupted time series studies (published between 2013 and 2017) found that graphs often do not meet core graphing recommendations.

MultiCenter Interrupted Time Series Analysis: Incorporating Within and Between-Center Heterogeneity

Background

Segmented regression (SR) is the most common statistical method used in the analysis of interrupted time series (ITS) data. However, this modeling strategy is indicated to produce spurious results when applied to aggregated data. For multicenter ITS studies, data at a given time point are often aggregated across different participants and settings; thus, conventional segmented regression analysis may not be an optimal approach. Our objective is to provide a robust method for analysis of ITS data, while accounting for two sources of heterogeneity, between participants and across sites.

Methods

We present a methodological framework within the segmented regression modeling strategy, where we introduced weights to account for between-participant variation and the differences across multiple sites. We empirically compared the proposed weighted segmented regression (wSR) with the conventional SR as well as with a previously published pooled analysis method using data from the Mobility of Vulnerable Elders in Ontario (MOVE-ON) project, a multisite ITS study.

Results

Overall, the wSR produced the most precise estimates, where they had the narrowest 95% CI, while the conventional SR method resulted in the least precise estimates. Our method also resulted in increased power. The pooled analysis method and the wSR had comparable results when there were ≤4 sites included in the overall analysis and when there was moderate to high between-site heterogeneity as measured by the I² statistic.

Conclusion

Incorporating participant-level and site-level variability led to estimates that were more precise and accurate in determining the magnitude of the effect of an intervention and led to increased statistical power. This underscores the importance of accounting for the inherent variability in aggregated data. Extensive simulations are required to further assess the methods in a wide range of scenarios and outcome types.

Methods, Applications and Challenges in the Analysis of Interrupted Time Series Data: A Scoping Review

Objective

Interrupted time series (ITS) designs are robust quasi-experimental designs commonly used to evaluate the impact of interventions and programs implemented in healthcare settings. This scoping review aims to 1) identify and summarize existing methods used in the analysis of ITS studies conducted in health research, 2) elucidate their strengths and limitations, 3) describe their applications in health research and 4) identify any methodological gaps and challenges.

Design

Scoping review.

Data Sources

Searches were conducted in MEDLINE, JSTOR, PUBMED, EMBASE, CINAHL, Web of Science and the Cochrane Library from inception until September 2017.

Study Selection

Studies in health research involving ITS methods or reporting on the application of ITS designs.

Data Extraction

Screening of studies was completed independently and in duplicate by two reviewers. One reviewer extracted the data from relevant studies in consultations with a second reviewer. Results of the review were presented with respect to methodological and application areas, and data were summarized using descriptive statistics.

Results

A total of 1389 articles were included, of which 98.27% (N=1365) were application papers. Segmented linear regression was the most commonly used method (26%, N=360). A small percentage (1.73%, N=24) were methods papers, of which 11 described either the development of novel methods or improvement of existing methods, 7 adapted methods from other areas of statistics, while 6 provided comparative assessment of conventional ITS methods.

Conclusion

A significantly increasing trend in ITS use over time is observed, where its application in health research almost tripled within the last decade. Several statistical methods are available for analyzing ITS data. Researchers should consider the types of data and validate the required assumptions for the various methods. There is a significant methodological gap in ITS analysis involving aggregated data, where analyses involving such data did not account for heterogeneity across patients and hospital settings.

What is the impact of psychiatric decision units on mental health crisis care pathways? Protocol for an interrupted time series analysis with a synthetic control study

BACKGROUND

The UK mental health system is stretched to breaking point. Individuals presenting with mental health problems wait longer at the ED than those presenting with physical concerns and finding a bed when needed is difficult - 91% of psychiatric wards are operating at above the recommended occupancy rate. To address the pressure, a new type of facility - psychiatric decision units (also known as mental health decision units) - have been introduced in some areas. These are short-stay facilities, available upon referral, targeted to help individuals who may be able to avoid an inpatient admission or lengthy ED visit. To advance knowledge about the effectiveness of this service for this purpose, we will examine the effect of the service on the mental health crisis care pathway over a 4-year time period; the 2 years proceeding and following the introduction of the service. We use aggregate service level data of key indicators of the performance of this pathway.

METHODS

Data from four mental health Trusts in England will be analysed using an interrupted time series (ITS) design with the primary outcomes of the rate of (i) ED psychiatric presentations and (ii) voluntary admissions to mental health wards. This will be supplemented with a synthetic control study with the same primary outcomes, in which a comparable control group is generated for each outcome using a donor pool of suitable National Health Service Trusts in England. The methods are well suited to an evaluation of an intervention at a service delivery level targeting population-level health outcome and the randomisation or 'trialability' of the intervention is limited. The synthetic control study controls for national trends over time, increasing our confidence in the results. The study has been designed and will be carried out with the involvement of service users and carers.

DISCUSSION

This will be the first formal evaluation of psychiatric decision units in England. The analysis will provide estimates of the effect of the decision units on a number of important service use indicators, providing much-needed information for those designing service pathways.

TRIAL REGISTRATION

primary registry: isrctn.com Identifying number: ISRCTN77588384 Link: Date of registration in primary registry: 27/02/2020.

PRIMARY SPONSOR

St George's, University of London, Cramner Road, Tooting, SW17 ORE. Primary contact: Joe Montebello.

Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: a review

OBJECTIVES

Interrupted time series (ITS) designs are frequently used in public health to examine whether an intervention or exposure has influenced health outcomes. Few reviews have been undertaken to examine the design characteristics, statistical methods, and completeness of reporting of published ITS studies.

STUDY DESIGN AND SETTING

We used stratified random sampling to identify 200 ITS studies that evaluated public health interventions or exposures from PubMed (2013-2017). Study characteristics, details of statistical models and estimation methods used, effect metrics, and parameter estimates were extracted. From the 200 studies, 230 time series were examined.

RESULTS

Common statistical methods used were linear regression (31%, 72/230) and autoregressive integrated moving average (19%, 43/230). In 17% (40/230) of the series, we could not determine the statistical method used. Autocorrelation was acknowledged in 63% (145/230) of the series. An estimate of the autocorrelation coefficient was given for only 1% of the series (3/230). Measures of precision were reported for 63% of effect measures (541/852).

CONCLUSION

Many aspects of the design, methods, analysis, and reporting of ITS studies can be improved, particularly description of the statistical methods and approaches to adjust for and estimate autocorrelation. More guidance on the conduct and reporting of ITS studies is needed to improve this study design.

Implementation of the Targeted Assessment for Prevention Strategy in a healthcare system to reduce Clostridioides difficile infection rates

BACKGROUND

Prevention of Clostridioides difficile infection (CDI) is a national priority and may be facilitated by deployment of the Targeted Assessment for Prevention (TAP) Strategy, a quality improvement framework providing a focused approach to infection prevention. This article describes the process and outcomes of TAP Strategy implementation for CDI prevention in a healthcare system.

METHODS

Hospital A was identified based on CDI surveillance data indicating an excess burden of infections above the national goal; hospitals B and C participated as part of systemwide deployment. TAP facility assessments were administered to staff to identify infection control gaps and inform CDI prevention interventions. Retrospective analysis was performed using negative-binomial, interrupted time series (ITS) regression to assess overall effect of targeted CDI prevention efforts. Analysis included hospital-onset, laboratory-identified C. difficile event data for 18 months before and after implementation of the TAP facility assessments.

RESULTS

The systemwide monthly CDI rate significantly decreased at the intervention (β2, -44%; P = .017), and the postintervention CDI rate trend showed a sustained decrease (β1 + β3; -12% per month; P = .008). At an individual hospital level, the CDI rate trend significantly decreased in the postintervention period at hospital A only (β1 + β3, -26% per month; P = .003).

CONCLUSIONS

This project demonstrates TAP Strategy implementation in a healthcare system, yielding significant decrease in the laboratory-identified C. difficile rate trend in the postintervention period at the system level and in hospital A. This project highlights the potential benefit of directing prevention efforts to facilities with the highest burden of excess infections to more efficiently reduce CDI rates.

Experimental and quasi-experimental designs in implementation research

Implementation science is focused on maximizing the adoption, appropriate use, and sustainability of effective clinical practices in real world clinical settings. Many implementation science questions can be feasibly answered by fully experimental designs, typically in the form of randomized controlled trials (RCTs). Implementation-focused RCTs, however, usually differ from traditional efficacy- or effectiveness-oriented RCTs on key parameters. Other implementation science questions are more suited to quasi-experimental designs, which are intended to estimate the effect of an intervention in the absence of randomization. These designs include pre-post designs with a non-equivalent control group, interrupted time series (ITS), and stepped wedges, the last of which require all participants to receive the intervention, but in a staggered fashion. In this article we review the use of experimental designs in implementation science, including recent methodological advances for implementation studies. We also review the use of quasi-experimental designs in implementation science, and discuss the strengths and weaknesses of these approaches. This article is therefore meant to be a practical guide for researchers who are interested in selecting the most appropriate study design to answer relevant implementation science questions, and thereby increase the rate at which effective clinical practices are adopted, spread, and sustained.

Improvement of perioperative care of the elderly patient (PeriAge): protocol of a controlled interventional feasibility study

INTRODUCTION

Geriatric patients have a pronounced risk to suffer from postoperative complications. While effective risk-specific perioperative measures have been studied in controlled experimental settings, they are rarely found in routine healthcare. This study aims (1) to implement a multicomponent preoperative and intraoperative intervention, and investigate its feasibility, and (2) exploratorily assess the effectiveness of the intervention in routine healthcare.

METHODS AND ANALYSIS

Feasibility and exploratory effectiveness of the intervention will be investigated in a monocentric, prospective, non-randomised, controlled trial. The intervention includes systematic information for patients and family about measures to prevent postoperative complications; preoperative screening for frailty, malnutrition, strength and mobility with nutrient supplementation and physical exercise (prehabilitation) as needed. Further components focus on potentially inadequate medication, patient blood-management and carbohydrate loading prior to surgery, retainment of orientation aids in the operating room and a geriatric anaesthesia concept. Data will successively be collected from control, implementation and intervention groups. Patients aged 65+ with impending surgery will be included. A sample size of 240, n=80 per group, is planned. Assessments will take place at inclusion and 2, 30 and 180 days after surgery. Mixed-methods analyses will be performed. Exploratory effectiveness will be assessed using mixed segmented regressions. The primary endpoint is functional status. Secondary endpoints include cognitive performance, health-related quality of life, length of inpatient stay and occurrence of postoperative complications. Feasibility will be assessed through semi-structured interviews with staff and patients and quantitative analyses of the data quality, focussing on practicability, acceptance, adoption and fidelity to protocol.

ETHICS AND DISSEMINATION

The study will be carried out in accordance with the Helsinki Declaration and to principles of good scientific practice. The Ethics Committee of the Medical Association Hamburg, Germany, approved the protocol (study ID: PV5596). Results will be disseminated in scientific journals and healthcare conferences.

TRIAL REGISTRATION NUMBER

ClinicalTrials.gov Identifier: NCT03325413.

Australian state influenza notifications and school holiday closures in 2019

Background: The impact of school holidays on influenza rates has been sparsely documented in Australia. In 2019, the early winter influenza season coincided with mid-year school breaks, enabling us the unusual opportunity to examine how influenza incidence changed during school holiday closure dates. Methods: The weekly influenza data from five Australian state and one territory health departments for the period of week 19 (mid-May) to week 39 (early October) 2019 were compared to each state's public-school holiday closure dates. We used segmented regression to model the weekly counts and a negative binomial distribution to account for overdispersion due to autocorrelation. The models' goodness-of-fit was assessed by plots of observed versus expected counts, plots of residuals versus predicted values, and Pearson's Chi-square test. The main exposure was the July two-week school holiday period, using a lag of one week. The effect is estimated as a percent change in incidence level, and in slope. Results: School holidays were associated with significant declines in influenza incidence in three states and one territory by between 41% and 65%. Two states did not show evidence of declines although one of those states had already passed its peak by the time of the school holidays. The models showed acceptable goodness-of-fit. The first decline during school holidays is seen in the school aged (5-19 years) population, with the declines in the adult and infant populations being smaller and following a week later. Conclusions: Given the significant and rapid reductions in incidence, these results have important public health implications. Closure or extension of holiday periods could be an emergency option for state governments.

Reduction in Clostridium difficile infection rates following a multifacility prevention initiative in Orange County, California: A controlled interrupted time series evaluation

OBJECTIVE

To evaluate the Orange County Clostridium difficile infection (CDI) prevention collaborative's effect on rates of CDI in acute-care hospitals (ACHs) in Orange County, California.

DESIGN

Controlled interrupted time series.

METHODS

We convened a CDI prevention collaborative with healthcare facilities in Orange County to reduce CDI incidence in the region. Collaborative participants received onsite infection control and antimicrobial stewardship assessments, interactive learning and discussion sessions, and an interfacility transfer communication improvement initiative during June 2015-June 2016. We used segmented regression to evaluate changes in monthly hospital-onset (HO) and community-onset (CO) CDI rates for ACHs. The baseline period comprised 17 months (January 2014-June 2015) and the follow-up period comprised 28 months (September 2015-December 2017). All 25 Orange County ACHs were included in the CO-CDI model to account for direct and indirect effects of the collaborative. For comparison, we assessed HO-CDI and CO-CDI rates among 27 ACHs in 3 San Francisco Bay Area counties.

RESULTS

HO-CDI rates in the 15 participating Orange County ACHs decreased 4% per month (incidence rate ratio [IRR], 0.96; 95% CI, 0.95-0.97; P &lt; .0001) during the follow-up period compared with the baseline period and 3% (IRR, 0.97; 95% CI, 0.95-0.99; P = .002) per month compared to the San Francisco Bay Area nonparticipant ACHs. Orange County CO-CDI rates declined 2% per month (IRR, 0.98; 95% CI, 0.96-1.00; P = .03) between the baseline and follow-up periods. This decline was not statistically different from the San Francisco Bay Area ACHs (IRR, 0.97; 95% CI, 0.95-1.00; P = .09).

CONCLUSIONS

Our analysis of ACHs in Orange County provides evidence that coordinated, regional multifacility initiatives can reduce CDI incidence.

Are national suicide prevention programs effective? A comparison of 4 verum and 4 control countries over 30 years

BACKGROUND

Suicide and non-fatal suicidal behavior are significant public health issues worldwide requiring effective preventive interventions.

METHODS

The aim of the present study was to analyze the effectiveness of national suicide prevention programs taking a statistical approach involving the segmented regression analysis of interrupted time series data.

RESULTS

This study demonstrates that National Suicide Prevention Programs are effective, but this effect seems to correlate with age and sex. Our data have shown a statistical significant decline in suicide rates in the verum countries in males, with the strongest effects in groups aged 25-to-44 years and 45-to-64 years.

CONCLUSION

Our study implies that the implementation of a national strategy is an effective tool to reduce suicide rates.

The value of the infection prevention and control nurse led MRSA ward round

Meticillin-resistant S. aureus (MRSA) is prevalent in most parts of the world. The study took place at Queen Elizabeth Hospital Birmingham (QEHB) a UK tertiary referral hospital. At QEHB innovative nurse led daily ward rounds for patients that acquire hospital acquired MRSA during their hospital stay are undertaken. The aim is to optimise care delivered for these patients whilst at QEHB, thereby reducing the risk of infection in patients with healthcare-acquired MRSA. A segmented Poisson regression model suggests that the MRSA bacteraemia rate was affected where an 88.94% reduction (p = 0.0561) in bacteraemias was seen by the introduction of these ward rounds. We describe a nurse led MRSA ward round which was associated with a lower rate of MRSA bacteraemias.

Control of Carbapenem-resistant Enterobacteriaceae, Acinetobacter baumannii, and Pseudomonas aeruginosa in Healthcare Facilities: A Systematic Review and Reanalysis of Quasi-experimental Studies

Carbapenem-resistant Enterobacteriaceae (CRE), Acinetobacter baumannii (CRAB), and Pseudomonas aeruginosa (CRPsA) are a serious cause of healthcare-associated infections, although the evidence for their control remains uncertain. We conducted a systematic review and reanalysis to assess infection prevention and control (IPC) interventions on CRE-CRAB-CRPsA in inpatient healthcare facilities to inform World Health Organization guidelines. Six major databases and conference abstracts were searched. Before-and-after studies were reanalyzed as interrupted time series if possible. Effective practice and organization of care (EPOC) quality criteria were used. Seventy-six studies were identified, of which 17 (22%) were EPOC-compatible and interrupted time series analyses, assessing CRE (n = 11; 65%), CRAB (n = 5; 29%) and CRPsA (n = 3; 18%). IPC measures were often implemented using a multimodal approach (CRE: 10/11; CRAB: 4/5; CRPsA: 3/3). Among all CRE-CRAB-CRPsA EPOC studies, the most frequent intervention components included contact precautions (90%), active surveillance cultures (80%), monitoring, audit and feedback of measures (80%), patient isolation or cohorting (70%), hand hygiene (50%), and environmental cleaning (40%); nearly all studies with these interventions reported a significant reduction in slope and/or level. The quality of EPOC studies was very low to low.

Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: protocol for a review

INTRODUCTION

An interrupted time series (ITS) design is an important observational design used to examine the effects of an intervention or exposure. This design has particular utility in public health where it may be impracticable or infeasible to use a randomised trial to evaluate health system-wide policies, or examine the impact of exposures (such as earthquakes). There have been relatively few studies examining the design characteristics and statistical methods used to analyse ITS designs. Further, there is a lack of guidance to inform the design and analysis of ITS studies.This is the first study in a larger project that aims to provide tools and guidance for researchers in the design and analysis of ITS studies. The objectives of this study are to (1) examine and report the design characteristics and statistical methods used in a random sample of contemporary ITS studies examining public health interventions or exposures that impact on health-related outcomes, and (2) create a repository of time series data extracted from ITS studies. Results from this study will inform the remainder of the project which will investigate the performance of a range of commonly used statistical methods, and create a repository of input parameters required for sample size calculation.

METHODS AND ANALYSIS

We will collate 200 ITS studies evaluating public health interventions or the impact of exposures. ITS studies will be identified from a search of the bibliometric database PubMed between the years 2013 and 2017, combined with stratified random sampling. From eligible studies, we will extract study characteristics, details of the statistical models and estimation methods, effect metrics and parameter estimates. Further, we will extract the time series data when available. We will use systematic review methods in the screening, application of inclusion and exclusion criteria, and extraction of data. Descriptive statistics will be used to summarise the data.

ETHICS AND DISSEMINATION

Ethics approval is not required since information will only be extracted from published studies. Dissemination of the results will be through peer-reviewed publications and presentations at conferences. A repository of data extracted from the published ITS studies will be made publicly available.

Wiping out MRSA: effect of introducing a universal disinfection wipe in a large UK teaching hospital

Background

Contamination of the inanimate environment around patients constitutes an important reservoir of MRSA. Here we describe the effect of introducing a universal disinfection wipe in all wards on the rates of MRSA acquisitions and bacteraemias across a large UK teaching hospital.

Methods

A segmented Poisson regression model was used to detect any significant changes in the monthly numbers per 100,000 bed days of MRSA acquisitions and bacteraemias from April 2013 - December 2017 across QEHB.

Results

From April 2013 to April 2016, cleaning of ward areas and multi-use patient equipment by nursing staff consisted of a two-wipe system. Firstly, a detergent wipe was used, which was followed by a disinfection step using an alcohol wipe. In May 2016, QEHB discontinued the use of a two-wipe system for cleaning and changed to a one wipe system utilising a combined cleaning and disinfection wipe containing a quaternary ammonium compound. The segmented Poisson regression model demonstrated that the rate of MRSA acquisition/100,000 patient bed days was affected by the introduction of the new wiping regime (20.7 to 9.4 per 100,000 patient bed days; p <0.005).

Discussion

Using a Poisson model we demonstrated that the average hospital acquisition rate of MRSA/100,000 patient bed days reduced by 6.3% per month after the introduction of the new universal wipe.

Conclusion

We suggest that using a simple one wipe system for nurse cleaning is an effective strategy to reduce the spread and incidence of healthcare associated MRSA.

Tap out: reducing waterborne Pseudomonas aeruginosa transmission in an intensive care unit

BACKGROUND

Pseudomonas aeruginosa is a ubiquitous and important opportunistic pathogen in immunocompromised or critically ill patients. Nosocomial P. aeruginosa outbreaks have been associated with hospital water sources.

AIM

To describe engineering interventions to minimize contamination of water outlets and the subsequent clinical impact.

METHODS

New tap outlets were fitted at selected outlets across the intensive care unit (ICU). Laboratory testing demonstrated that, following artificial contamination with P. aeruginosa, these taps could be effectively decontaminated using a thermal washer-disinfector. Water samples were collected weekly from new outlets on the ICU over an eight-month period and tested for the enumeration of P. aeruginosa via membrane filtration. Surveillance of P. aeruginosa from clinical specimens was routinely undertaken.

FINDINGS

Prior to the interventions, water sampling on ICU indicated that 30% of the outlets were positive for P. aeruginosa at any one time, and whole genome sequencing data suggested at least 30% transmission from water to patient. Since their installation, weekly sampling of the new tap outlets has been negative for P. aeruginosa, and the number of P. aeruginosa clinical isolates has fallen by 50%.

CONCLUSION

Installation and maintenance of tap outlets free of P. aeruginosa can substantially reduce the number of P. aeruginosa clinical isolates in an ICU.

Methods, applications, interpretations and challenges of interrupted time series (ITS) data: protocol for a scoping review

OBJECTIVES

Interrupted time series (ITS) design involves collecting data across multiple time points before and after the implementation of an intervention to assess the effect of the intervention on an outcome. ITS designs have become increasingly common in recent times with frequent use in assessing impact of evidence implementation interventions. Several statistical methods are currently available for analysing data from ITS designs; however, there is a lack of guidance on which methods are optimal for different data types and on their implications in interpreting results. Our objective is to conduct a scoping review of existing methods for analysing ITS data, to summarise their characteristics and properties, as well as to examine how the results are reported. We also aim to identify gaps and methodological deficiencies.

METHODS AND ANALYSIS

We will search electronic databases from inception until August 2016 (eg, MEDLINE and JSTOR). Two reviewers will independently screen titles, abstracts and full-text articles and complete the data abstraction. The anticipated outcome will be a summarised description of all the methods that have been used in analysing ITS data in health research, how those methods were applied, their strengths and limitations and the transparency of interpretation/reporting of the results. We will provide summary tables of the characteristics of the included studies. We will also describe the similarities and differences of the various methods.

ETHICS AND DISSEMINATION

Ethical approval is not required for this study since we are just considering the methods used in the analysis and there will not be identifiable patient data. Results will be disseminated through open access peer-reviewed publications.

Universal Mask Usage for Reduction of Respiratory Viral Infections After Stem Cell Transplant: A Prospective Trial

Background. Respiratory viral infections (RVIs) are frequent complications of hematopoietic stem cell transplant (HSCT). Surgical masks are a simple and inexpensive intervention that may reduce nosocomial spread.

Methods. In this prospective single-center study, we instituted a universal surgical mask policy requiring all individuals with direct contact with HSCT patients to wear a surgical mask, regardless of symptoms or season. The primary endpoint was the incidence of RVIs in the mask period (2010–2014) compared with the premask period (2003–2009).

Results. RVIs decreased from 10.3% (95/920 patients) in the premask period to 4.4% (40/911) in the mask period (P < .001). Significant decreases occurred after both allogeneic (64/378 [16.9%] to 24/289 [8.3%], P = .001) and autologous (31/542 [5.7%] to 16/622 [2.6%], P = .007) transplants. After adjusting for multiple covariates including season and year in a segmented longitudinal analysis, the decrease in RVIs remained significant, with risk of RVI of 0.4 in patients in the mask group compared with the premask group (0.19–0.85, P = .02). In contrast, no decrease was observed during this same period in an adjacent hematologic malignancy unit, which followed the same infection control practices except for the mask policy. The majority of this decrease was in parainfluenza virus 3 (PIV3) (8.3% to 2.2%, P < .001).

Conclusions. Requiring all individuals with direct patient contact to wear a surgical mask is associated with a reduction in RVIs, particularly PIV3, during the most vulnerable period following HSCT.