Interrupted Time Series Research Articles

Larval source management in Ethiopia: modelling to assess its effectiveness in curbing malaria surge in dire Dawa and Batu Towns

BackgroundEthiopia faces several severe challenges in terms of malaria elimination, including drug resistance and diagnostic evasion in the Plasmodium falciparum parasite, insecticide resistance in the primary Anopheles malaria vector, and, most recently, the invasion of the Asian malaria vector Anopheles stephensi. Novel malaria control methods are therefore needed, and in this paper, we describe the evaluation of a larval source management (LSM) strategy implemented in response to An. stephensi. The primary outcome was the malaria incidence rate compared between intervention and non-intervention sites in the presence of An. stephensi.MethodsIntervention (Batu and Dire Dawa) and control (Metehara) towns were selected, and weekly malaria passive case detection data collected between 2014 and 2023 were obtained from the Oromia regional state and Dire Dawa City Administration Health Bureau. In addition, data regarding intervention were obtained from the President’s Malaria Initiative (PMI) reports. Weekly malaria passive case data were used to evaluate the change in the estimated malaria incidence rate and trends of temporal patterns of the estimated malaria incidence rate before and after interventions. An interrupted time series model with a cyclic second-order random walk structure periodic seasonal term was used to assess the impact of LSM on malaria incidence rate in the intervention and control settings.ResultsAn upsurge in malaria cases occurred after 2020 at both the intervention and control sites. The temporal patterns of malaria incidence rate showed an increasing trend after the intervention. The ITS model depicted that the LSM has no impact in reducing the malaria incidence rate at both intervention site Dire Dawa [immediate impact = 1.462 (0.891, 2.035)], [Lasting impact = 0.003 (− 0.012, 0.018)], and Batu [Immediate impact 0.007 (− 0.235, 0.249), [Lasting impact = 0.008 (− 0.003, 0.013)].ConclusionsAn overall increasing trend in the malaria incidence rate was observed irrespective of the implementation of LSM in the urban settings of Ethiopia, where An. stephensi has been found. Further investigations and validations of the incorporation of LSM into control activities are warranted.

Absenteeism Among National Health Service Workers During the COVID-19 Pandemic in Portugal

By March 2020, the first COVID-19 cases were detected in Portugal. Anecdotal evidence showed that the National Health Service absenteeism rate rose. The present study aimed to quantify and characterize the absenteeism rate among these healthcare workers during the COVID-19 pandemic in Portugal. This study used data from the Portuguese National Health Service Transparency Portal for the number of National Health Service workers and absence days (2015 - 2021). Absenteeism was compared, before and after the pandemic onset, in absolute terms, and as absenteeism rates. Additionally, we performed an interrupted time series analysis, by fitting a Poisson regression model with level change. We controlled for data seasonality using Fourier terms. The results showed that, between 2015 and March 2020, the average monthly absenteeism rate was of 12.2%, rising to 14.5% in the subsequent period. This represented an increase of 19% in the absenteeism rate. The interrupted time series analysis showed an increase of 10.8% in the National Health Service absenteeism rate after the pandemic onset [relative risk = 1.10; 95% CI 1.10 - 1.11; p < 0.01]. When accounting for seasonality in the data, the interrupted time series analysis showed an increase of 11.0% in the National Health Service absenteeism rate [relative risk = 1.11; 95% CI 1.00 - 1.23; p < 0.05]. These results suggest an increased number of days of absence rate among the National Health Service healthcare workers during the COVID-19 pandemic.

Analysing the effects of National Centralised Drug Procurement and Price Negotiation Policies on novel hypoglycaemic drug usage and costs in Shanghai, China: an interrupted time series analysis

ObjectivesTo analyse the impacts of multiple rounds of National Centralised Drug Procurement Policy (NCDP) and National Drug Price Negotiation Policy (NDPN) on the utilisation, expenditure and daily cost of novel...

Information provision and financial incentives in Catalonia's public primary care (2010–2019): an interrupted time series analysis

The relative efficacy of information provision versus financial incentives in improving primary care quality remains a critical, unresolved question. We investigated these two strategies in Catalonia's public primary care system from 2010 to 2019: an innovative online platform providing real-time quality indicator information and targeted economic incentives for achieving indicator goals. We conducted a comprehensive interrupted time series regression analysis on data from 272 primary care practices (5,628,080 patients). This analysis used linear regression models with Newey-West standard errors, and a sensitivity analysis including logit transformations to address ceiling effects. We evaluated 1) immediate post-intervention changes (step changes) in indicator results and inter-practice variability (coefficient of variation, CV), and 2) shifts in pre-intervention trends (slopes). We scrutinized 39 indicators after rigorous quality control: 23 novel (12 informed, 11 incentivized) and 16 derived from existing incentivized indicators. Robustness checks included 14 consistently incentivized and 10 non-intervened indicators. Overall, we assessed 63 indicators: 18 control, 13 follow-up, 9 quaternary prevention, 7 treatment, 7 diagnosis, 6 screening and 3 vaccination indicators. Informed indicators showed positive impacts in 75% (9/12) of cases, and incentivized indicators in 64% (7/11) of cases. Incentivized indicators displayed improvements in annual trends ranging from 6.66 to 1.25 percentage points, with step changes up to 8.87 percentage points. Information led to step changes ranging from 19.67 to 1.07 percentage points, along with trend improvements between 1.09 and 0.34 percentage points annually. Both interventions were associated with step reductions in variability (up to-0.18 CV reduction) and significant trend improvements. Derived indicators showed limited improvements in results or variability (31%, 5/16), with minor step increases up to 2.22 percentage points. Our findings reveal that information provision alone can match or even surpass the impact of financial incentives in improving care quality and reducing practice variability. This challenges conventional wisdom and offers a cost-effective, scalable approach to primary care quality enhancement, with far-reaching implications for global health policy. European Union, Horizon Europe.

The Impact of the COVID-19 Pandemic on Police Involvement in Mental Health Calls for Service

The onset of the COVID-19 pandemic not only threatened individuals’ physical health but also caused a serious strain on mental health and access to care. Analyzing 4 years of data from 2019 through 2022, this study examines whether the frequency of mental health calls for service and police-initiated stops for mental health reasons changed over the course of the COVID-19 pandemic. Using interrupted time-series analysis, the results show that police involvement in mental health care was significantly affected by the COVID-19 pandemic. However, different changes were observed between dispatched calls for service and police-initiated stops for mental health reasons.

Estimating the impact of the first 2 years of the COVID-19 pandemic on tuberculosis diagnosis and treatment outcomes in Southeastern City in Iran: an interrupted time series analysis of the preceding 10 years of ecological data

BackgroundWith shared modes of transmission and clinical symptoms the convergence of COVID-19 and tuberculosis (TB) might lead to reduced diagnosis and detection of TB, which is challenging for healthcare systems already strained by the pandemic's reach.MethodsThis ecological study investigated the impact of the COVID-19 pandemic on TB surveillance over the first 2 years of the pandemic (March 2020 to February 2022) in southeastern Iran. Interrupted Time Series (ITS) analysis with the quasi-Poisson regression models was used to estimate the relative risk (RR) of TB diagnosis and treatment outcome counts, stratified by gender, case definition, involvement type, and treatment outcomes.ResultsThe ITS analyses showed a significant decrease in TB total cases (RR: 0.622 [95% CI: 0.487, 0.793], P < 0.001), new cases (RR: 0.632 [95% CI: 0.493, 0.810], P < 0.001) and recurrent cases (RR: 0.491 [95% CI: 0.247, 0.974], P < 0.001). In addition, recovery and treatment failure counts also showed significant decreases (RR: 0.751 [95% CI: 0.566, 0.996], P = 0.05; RR: 0.201 [95% CI: 0.054, 0.738], P = 0.02). Moreover, significant decreases are observed in both genders and involvement types (pulmonary and extrapulmonary TB). No significant change was observed for absent to treatment and death counts.ConclusionThe COVID-19 pandemic has negatively impacted TB diagnosis and treatments. Concerns are risen about the progress achieved in TB control.

Segregation and school composition in Washington, D.C. following the implementation of a unified school lottery

ABSTRACT The global expansion of choice programmes fuels debates about the equity potential of school choice policies and mechanisms. In 2014, Washington, D.C. pioneered a citywide lottery and enrolment system to match students with preferred institutions and distribute access to seats in high-demand schools. While unified lottery and enrolment systems impact a vast number of students and influence educational opportunities across continents, little evidence exists on the efficacy of these systems or the link between centralized lotteries and changes in school composition or segregation. Using longitudinal data and an interrupted time series analysis, this study examines the relationship between D.C.’s centralized lottery system and student distribution. We find unchanged desegregation trends systemwide and annual increases in segregation in schools located in wealthier and lower-income wards after the adoption of the lottery system. We raise questions about the way school-designated preference structures give the appearance of numerous school options for families, while keeping access to meaningful choices superficial. Across international contexts, these systems may aid desegregation efforts and enhance equity by randomizing school matches, preventing cream skimming, and offering a baseline of school information, but additional research is needed to better understand their substantive outcomes.

Remote school instruction in Fall 2020 and psychiatric emergencies among adolescents in Los Angeles County

ObjectiveSchools play an essential role in providing mental health care for adolescents. School closures during COVID-19, as well as re-opening to remote-only instruction in Fall 2020, may indirectly affect the utilization of emergency psychiatric care. We examine COVID-19-related changes in emergency psychiatric care among youth during the school closure and after school reopening (with remote instruction).MethodsWe use Box-Jenkins interrupted time series methods to analyze psychiatric emergency department (ED) visits among patients 10–19 years at LAC + USC Medical Center (LAC + USC) between January 5th, 2018, and December 31st, 2020. We control for the 1st societal shutdown in LA County (i.e., the nine weeks from March 13 to May 14, 2020) when analyzing the potential “return to remote school” shock.ResultsYouth psychiatric ED visits fell by 15.3 per week during the Spring 2020 school closure (p < .05). The “return to remote school” coefficient (i.e., August 14th to September 10, 2020), by contrast, is positive but does not reach statistical detection above expected values (p = .11). However, the proportion of psychiatric ED visits rises 38% among youth during the “return to remote school” period (p = 0.006).ConclusionThe initiation of Fall 2020 remote instruction corresponded with a greater proportion of youth ED visits that are classified as psychiatric.

Impact of Revised Diagnostic Testing Algorithms and Case Definitions on Lyme Disease Surveillance in a Low-Incidence State: An Interrupted Time Series Analysis.

We evaluated the impact of the endorsement of modified 2-tiered testing and its inclusion in updated Lyme disease case definitions in North Carolina using surveillance data from 2017-2023. After the implementation of modified 2-tiered testing, confirmed cases increased by 117%, with comparable decreases in probable cases compared with prior periods.

Inhalation corticosteroids for COVID-19 - a real world data analysis on guideline adherence.

The recommendation to consider prescribing inhalation corticosteroids to a subgroup of vulnerable COVID-19 patients was added to the Dutch medical guideline on November 11, 2021, and was also adopted by other countries during the pandemic. To evaluate the adherence of general practitioners to this guideline, and whether real-world data quality is sufficient to study the effect of revised guidelines on prescribing behaviour. A retrospective cohort study using Dutch primary care data from the Extramural LUMC Academic Network database, containing patient data of 129 general practices in the Leiden - The Hague area. We performed an interrupted time series analysis to measure the effect of the new guideline on the prescription rate of ICS, accounting for general trend and seasonal fluctuations. Between July 1, 2020 to August 1, 2022, 131,482 patients had 164,098 COVID-19 consultations. During this period, 1,709 patients received 2094 ICS prescriptions for COVID-19. After the guideline update, there was an instantaneous decrease in prescription rate (IRR 0.47, 95% CI 0.32-0.69). Prescription rate in the subgroup of vulnerable patients did not change significantly (IRR 0.93, 95% CI 0.66-1.32), while less vulnerable patients were significantly prescribed less (IRR 0.29, 95% CI 0.14-0.59). The revision to COVID-19 guidelines had significant impact on general practitioners' prescription behaviour soon after publication: prescription rate remained constant for vulnerable patients, while less vulnerable patient were significantly prescribed less often. Using electronic health records it is feasible to assess changes in guideline adherence using interrupted time series.

National impact of a constraining regulatory framework on pregabalin dispensations in France, 2020–2022

BackgroundPregabalin is a drug approved for neuropathic pain, epilepsy and general anxiety disorder. However, pregabalin is also an increasing cause of diversion and misuse, and, for this reason, the French health authorities have decided in 2021 to classify it as a narcotic drug, requiring secured prescription pads. Our study aimed to evaluate the impact of this measure on pregabalin dispensation patterns. MethodsUsing data from a national representative sample of 12,690 French community pharmacies, we assessed the impact of the new regulatory framework, implemented in June 2021, on the total monthly quantity of pregabalin dispensed by conducting an interrupted time-series (ITS) analysis. We built a counterfactual model to predict what would have happened without the intervention. We conducted the same analyses in the subsample of pregabalin dispensations exceeding the maximum recommended dosage of 600 mg per day. Additionally, we compared (1) the number of pregabalin dispensations (2) the co-dispensing of opioids and/or benzodiazepines one year before and one year after the regulation. ResultsFollowing the regulatory change, there was an immediate and significant drop of 38,475,375 mg (95 %CI [-66,931,799 mg; -10,018,951 mg] in pregabalin dispensations. This decline continued in the following months, with a decrease of 4,788,107 mg (95 %CI: [-8,888,326 mg; -687,888 mg] per month. The counterfactual model predicted that, without the intervention, there would have been no notable change in pregabalin dispensations. The ITS analyses showed no significant change in pregabalin dispensations following the new regulation for the subsample with doses exceeding 600 mg per day. Concurrently, there was a significant decrease in co-dispensing of opioids (18.4 % pre-regulation versus 11.6 % post-regulation, p < 0.001) and benzodiazepines (21.4 % pre-regulation versus 11.7 % post-regulation, p < 0.001). ConclusionThe constraining regulation adopted in 2021 by the French health authorities significantly curtailed total pregabalin dispensing in community pharmacies. Our findings underscore the importance of regulatory measures to limit the misuse of a prescribed medication.

Changes in food and drink purchasing behaviour in England during the first three months of the COVID-19 pandemic: An interrupted time series analysis.

This study examined changes food and drink purchasing during the first three months of the COVID-19 pandemic in England, and if changes varied by population subgroups. We investigated changes in take-home food and drink purchasing and frequency of out-of-home purchasing using an interrupted time series analysis design. The start of pandemic restrictions (the intervention) was defined as 16th March 2020, when first announced in the UK. London and the North of England. 1,245 households reporting take-home and 226 individuals reporting out-of-home purchases between January 2019 and mid-June 2020 from the GB Kantar Fast Moving Consumer Goods Panel. The marginal mean estimate of total take-home energy purchased was 17.4% (95% CI 14.9, 19.9) higher during the pandemic restriction period compared to the counterfactual. Increases of 35.2% (95% CI 23.4, 47.0) in take-home volume of alcoholic beverages and 1.2% (95% CI 0.1, 2.4) in foods and drinks high in fat, salt and sugar were observed. Reductions in purchased energy from fruit and vegetables (-7.3%, 95% CI -10.9, -3.6), ultra-processed foods (-4.0%, 95% CI -5.2, -2.8), and in out-of-home purchasing frequency (-44.0%, 95% CI -58.3, -29.6) were observed. Changes in chocolate and confectionery, soft drink and savoury snack purchases levelled off over time. Changes in all studied outcomes varied by sociodemographic characteristics and usual purchasing. Pandemic restrictions were associated with positive and negative changes in food and drink purchasing, which differed by individual characteristics. Future research should ascertain if changes persist and translate into changes in health.

Musculoskeletal Injury Surveillance inthe U.S. Military: Analysis ofthe ICD-10-CM Transition andDescriptive Report ofHealth Care Utilization at Four Sites Adopting MHS GENESIS.

The U.S. Military Health System (MHS) transitioned to the ICD-10-CM coding scheme in October 2015 and began the phased rollout of a new electronic health record system, MHS GENESIS, in October 2017. Both changes have the potential to affect the observed prevalence and health care utilization associated with musculoskeletal injuries (MSKIs) in service members. The purpose of this article was to (1) determine the effect of the ICD-10-CM transition on the observed prevalence of select MSKI conditions and (2) describe MSKI-related health care utilization in four MTFs during the adoption of MHS GENESIS. We calculated monthly prevalence rates for six diagnostic groupings of MSKIs routinely seen in the MHS between October 2011 and February 2020. To determine if the transition to ICD-10-CM influenced prevalence rates, we performed an interrupted time series analysis, using the ICD-10-CM transition date (October 1, 2015) as the interruption point. To assess trends in direct and private sector care encounters during the MHS GENESIS transition, we calculated monthly MSKI-related encounters at four MTFs from November 1, 2015 through September 30, 2021. Three diagnoses had a significant (P < .05) change in the slope, or rate of change, for their monthly prevalence after the introduction of ICD-10-CM (patellofemoral pain syndrome, -0.18; stress fractures, 0.12; subacromial pain, 0.03). These diagnoses also had a significant level change or immediate change in monthly prevalence following the ICD-10-CM transition (patellofemoral pain syndrome, 24.2; stress fractures, 0.16; subacromial pain, 0.36). Three of the four sites adopting MHS GENESIS showed reduced 3-month averages (range: -7.1-13.0%) of MSKI-related encounters following the electronic health record transition. For two of the four MTFs, we observed an increased reliance on private sector immediately leading up to and following the change to MHS GENESIS. The observed differences in monthly prevalence for certain MSKIs could be because of the changes in the amount and/or specificity of available codes in ICD-10-CM. Within the six selected diagnostic groupings of MSKIs, we found that patellofemoral pain syndrome, stress fractures, and subacromial pain syndrome demonstrated the greatest changes in prevalence during the ICD-10-CM transition. Those involved in MSKI surveillance should exercise caution when evaluating MSKI prevalence that spans the ICD-10-CM transition. Changes in health care utilization patterns in two of the four MTFs during their adoption of MHS GENESIS suggest that MSKI care may have transferred from direct to private sector care during this period. Future research will be required to validate this finding and determine the impacts on clinical outcomes and military readiness.

Impact of the Diagnosis-Intervention Packet Payment Reform on Provider Behavior in China: A Controlled Interrupted Time Series Study.

China has developed a novel case-based payment method called the DiagnosisIntervention Packet (DIP) to regulate healthcare providers' behavior. G city, a metropolis in southeast China, has shifted its payment policy from fixed rate per admission to DIP under regional global budget since 2018. This study examined the immediate and trend changes in provider behavior after this payment reform. Discharge data in G city between 2016 and 2019 was used, covering more than 10 million inpatient cases in 320 hospitals. A counterfactual scenario was developed to assign insured and uninsured inpatients across the study period to specific DIP groups under consistent rules. Controlled interrupted time-series (ITS) analyses were performed, with uninsured inpatients as control. Outcomes included inpatient volume, average DIP weight (similar to case-mix index in Diagnosis-Related Groups), and two innovative indicators (average diagnostic weight and average treatment weight) to decompose the changes in DIP weight. Subgroup analyses were conducted for different hospital levels and 21 major disease categories. After the DIP reform, monthly trend of inpatient volume decreased (-1085.34, P=0.052), while monthly growth of average DIP weight increased (2.17, P=0.02). No significant changes in average diagnostic weight were observed. Monthly trend of average treatment weight increased (2.38, P=0.001) after the reform. Secondary and tertiary hospitals experienced insignificantly decreased inpatient volume and elevated average DIP weight, accompanied by negligible change in average diagnostic weight and significant increase in average treatment weight. Primary hospitals experienced reduced inpatient volume and stable average DIP weight, along with increase in average diagnostic weight and decrease in average treatment weight. By differentiated payments for severity, DIP induced hospitals to shift their focus from volume to weight of inpatients. Instead of diagnostic upcoding, hospitals responded to the DIP reform primarily by increasing treatment intensity. Primary hospitals may face financial risks under regional competition.

Needs Assessment and Tailored Training Pilot for Emergency Care Clinicians in the Prehospital Setting in Rwanda

Background: In low- and middle-income countries (LMIC), 45% of deaths could be addressed by implementation of an emergency medical services (EMS) system. Prehospital care is a critical component of EMS worldwide, and basic, affordable training has been shown to improve EMS systems. However, patient outcome impact is unclear. In this study we aimed to assess the current state of prehospital care in Kigali, Rwanda, through a needs assessment, focused training intervention, and analysis of current practices and patient outcomes. Methods: We identified 30 clinicians through the prehospital medical command office and included them in the study. A prospective, nonrandomized, interrupted time-series approach was used. Data collected through closed- and open-ended questionnaires included age, sex, training, and knowledge assessment. We used the data to create a tailored, 18-hour training after which immediate and 11-month post-tests were administered. Linked prehospital and hospital care datasets allowed for evaluation of patient outcomes and prehospital process indicators that included training skill application, airway intervention, intravenous fluid administration, and glucose administration. Results: Of 30 clinicians, 18 (60%) were female, 19 were nurses, and 11 were nurse anaesthetists. Median age was 36, and median years providing care was 10 (IQR 7–11). Twenty-four (80%) participants completed immediate and post-test assessments. Mean knowledge across 12 core skills significantly improved from a pre-test mean of 59.7% (95% confidence interval [CI] 42.2–77.20) to a post-test mean of 87.8% (95% CI 74.7–100). At 11 months post-training, the score improvement maintained, with a mean score of 77.6%(95% CI 59.2–96.8). For patient outcomes, the total sample size was 572 patients; 324 of these patients were transported to the ED during the pre-training period (56.4%), while 248 were transported post-training. Prehospital oxygen administration for patients with a saturation level of &amp;lt;95% significantly increased pre- to post-training (66.7% to 71.7%; Δ = 5.0%; Δ95% CI 1.9,−8.1%). No significant changes were noted in patient treatment outcomes or other process indicators due to small sample sizes. Conclusion: This study provides insights on Rwandan EMS and demonstrates that a tailored intervention targeting education on prehospital process indicators has positive impacts on clinician knowledge and practice.

Trend of hand, foot and mouth disease before, during, and after China's COVID control policies in Zhejiang, China.

To describe the trends in the incidence of hand, foot and mouth disease (HFMD) before, during, and after China's Coronavirus Disease (COVID) control policies, and to interpret the influence on HFMD incidence at different control stages in Zhejiang Province. We collected data on HFMD cases in Zhejiang between 2014 and 2023. We compared the constituent ratios of cases at different COVID control stages by sex, age, child groups, and pathogens and weekly seasonal indices to observe seasonal variations in the incidence of HFMD. An interrupted time-series segmented regression analysis was applied to estimate the influence on HFMD incidence at different control stages. Stratified and sensitivity analyses were conducted to validate the findings. A considerable proportion of cases occurred among children living separately. The proportions of children in kindergartens or nurseries and children aged 2-4 years were relatively low at the strict control stage compared to the other three stages. Enteroviruses other than enterovirus 71 and coxsackie virus A16 were the dominant HFMD pathogens, and the proportion showed an increasing trend. The usual spring-summer peak in HMFD incidence did not occur in 2020, and the periodicity of the biennial peak was disrupted for a year. The summer peak in 2023 was higher than that in the other years, and was delayed by 3 weeks. The trend changes in weekly HFMD cases during the strict control and regular control stages were - 15% (IRR: 0.85, 95% CI: 0.81-0.89) and 17% (IRR: 1.17, 95% CI: 1.12-1.23), respectively. However, the change was not statistically significant during the reopening stage (IRR: 1.41, 95% CI: 0.34-5.88). The expected number of cases increased by 1.12 times (95% CI: 243.17, 53.45%) during the reopening stage compared to what would have occurred if the zero-COVID policy had continued in 2023. Non-pharmaceutical interventions (NPIs) for COVID-19 control can mitigate HFMD. However, after the dynamic zero-COVID policy ended, the HFMD incidence returned to historical levels. Strict NPIs such as traffic restrictions and kindergarten closures cannot be sustained long-term. NPIs such as improving personal hygiene for routine prevention are highly recommended.

Pediatric firearm injuries and socioeconomic vulnerability before and during the COVID-19 pandemic

BackgroundPediatric firearm injuries disproportionately affect groups experiencing socioeconomic disadvantage. Firearm injuries increased during the COVID-19 pandemic, but the impact on communities by degree of socioeconomic disadvantage is unknown. We examined the association between socioeconomic vulnerability and change in pediatric firearm injuries before versus during the pandemic. MethodsThis was a secondary analysis of a multicenter retrospective study comparing pediatric injuries pre- (March 17, 2019-December 31, 2020) versus during (March 15, 2020 – December 31, 2020) the COVID-19 pandemic. The parent study included injury-related visits to one of 40 Pediatric Emergency Departments in patients <18 years old. We examined firearm injuries as a proportion of all injuries. Deprivation Index (DI) was assigned using home zip code and divided into quartiles based on the distribution of DI in our total population, with deprivation increasing from quartile 1 (Q1, least deprivation) to quartile 4 (Q4, most deprivation). Kruskal-Wallis and Chi-square tests were used to compare variables of interest. Interrupted time series analysis was used to estimate pandemic effects on firearm injuries by DI quartile. ResultsThere were 1231 visits for firearm injuries. Mean DI was higher among firearm-injured than non-firearm-injured patients (0.43 vs. 0.35, p < 0.0001). Firearm injuries increased in proportion to all injuries pre- vs during pandemic (0.16 % vs 0.27 %, p < 0.0001). On interrupted time series analysis, the proportion of firearm injuries increased for Q2, Q3, and Q4. On subgroup analysis by intent, assault injuries as a proportion of all injuries increased from pre- to during COVID-19 for Q2 (3.32 % vs 6.94 %, ARD 3.62 %, 95 % CI 1.03, 6.20), Q3 (2.71 % vs 7.01 %, ARD 4.30, 95 % CI 1.99, 6.61), and Q4 (5.97 % vs 17.40 %; ARD 11.42, 95 % CI 0.19, 0.33). Unintentional injuries and injuries in youth 10–17 years old increased for Q2, Q3, and Q4. ConclusionsThe increase in pediatric firearm injuries during the COVID-19 pandemic impacted youth in all but the lowest deprivation quartile. Efforts at curbing gun violence should identify and amplify protective effects in under-resourced communities.

Estimating Mediation Effects in ABAB Reversal Designs.

Single-Case Experimental Designs (SCEDs), or N-of-1 trials, are commonly used to estimate intervention effects in many disciplines including in the treatment of youth mental health problems. SCEDs consist of repeated measurements of an outcome over time for a single case (e.g., student or patient) throughout one or more baseline phases and throughout one or more intervention phases. The manipulation of the baseline and intervention phase make the SCED a type of interrupted time series design, which is considered one of the most effective experimental designs for causal inference. An important step towards understanding why interventions are effective at producing a change in the outcome is through the investigation of mediating mechanisms. Hypotheses of mediating mechanisms involve an intervention variable which is hypothesized to affect an outcome through its effect on a mediating variable. Little work has attempted to combine mediation analysis and ABAB reversal designs. Therefore, the goals of this paper are to define, estimate, and interpret mediation effects for ABAB reversal designs. An empirical example is used to demonstrate how to estimate and interpret the mediation effects. R code is provided for researchers interested in estimating mediation effects in single-case reversal designs.

Self-poisoning in young Australians: The impact of COVID-19 and recent trends following easing of restrictions

Introduction: Self-poisoning is the most common form of hospitalised self-harm in young people and has been increasing over the last decade. The COVID-19 pandemic was associated with further increases in self-harm in adolescents. There are limited data on changes after pandemic restrictions ended. This study aimed to describe recent trends in self-poisoning in young people. Setting: Australia 2018–2023. Methods: A retrospective analysis of self-poisoning calls to Australia’s poisons information centres. Intentional poisoning exposures in people aged 5–19 years were included in the main analysis. We also examined intentional poisonings in people aged 20 and over for comparison. We used interrupted time-series analysis to estimate the impact of COVID-19 (March 2020) and ‘re-opening’ (December 2021), when many restrictions were lifted in Australia. We compared effects in the jurisdictions with most disruptions and lockdowns with other jurisdictions. Results: There were 201,755 self-poisonings documented, 64% female, with 19,946 children (aged 5–14, 9.9%) and 44,388 adolescents (aged 15–19, 22.0%). There was a steep increase in the rate of poisoning in female children and adolescents – with the annual increase being 1.361 (95% confidence interval = [1.074, 1.735]) and 1.154 (95% confidence interval = [1.049, 1.283]) times faster than pre-pandemic trends in these two groups between March 2020 and December 2021. Following relaxation of restrictions, there was a sudden drop and then slow decline in self-poisonings (despite rising COVID-19 infection rates). There was an overall 29.6% increase in intentional poisonings in young people over the study period. This is markedly more pronounced among young females (37.0%) vs males (11.2%). There were substantial jurisdictional variations, with larger increases in Victoria, Tasmania and the Australian Capital Territory. Conclusion: There has been a marked increase in child and adolescent female self-poisonings over the last 6 years, with increases strongly associated with COVID-19 pandemic restrictions but persisting after their removal.

Interrupted Time Series for Assessing the Causality of Intervention Effects. Part 35 of a Series on Evaluating Scientific Publications.

The gold standard for evaluating interventions in medicine and health care is the randomized controlled trial (RCT). In practice, however, RCTs may be difficult to conduct because of high costs, ethical aspects, or practical considerations. This is particularly true of studies on the population level, e.g., for the evaluation of health policy measures. We present a type of study design called the interrupted time series (ITS) and its critical interpretation, with several illustrative examples. This discussion is based on selected methodological publications. ITS are suitable for the assessment of interventions with a clear point of intervention (interruption). They are analyzed with the statistical methods of time-series analysis. One strength of ITS is that they can be used to estimate an immediate effect as well as a gradually developing effect. Under certain assumptions, the findings of an ITS analysis can be interpreted causally. The main assumption underlying an ITS is that the trend after the intervention would have been exactly the same as the trend before the intervention if the intervention had not taken place and all other conditions had remained unchanged. A further assumption is that there should be no differences in the pre- versus postintervention phases in the subjects or other entities being tested (e.g., hospitals) that might affect the measured endpoints (e.g., differences in mean age affecting measured mortality). Moreover, the intervention periods must be properly distinct from one another in order to prevent biased effect estimates. The robustness of the assumptions should also be checked with sensitivity analyses. As long as all sources of bias have been avoided and the findings are both plausible and robust, the effects revealed by ITS can be interpreted as causal. ITS may serve as an alternative method for evaluating intervention effects when an RCT cannot be performed.