scholarly journals Reflection on modern methods: a common error in the segmented regression parameterization of interrupted time-series analyses

2020 ◽  
Author(s):  
Hong Xiao ◽  
Orvalho Augusto ◽  
Bradley H Wagenaar
Keyword(s):  
Time Series ◽  
Randomized Study ◽  
Interrupted Time Series ◽  
Linear Segment ◽  
Segmented Regression ◽  
Calendar Time ◽  
Post Intervention ◽  
Common Error ◽  
Indicator Variable ◽  
Intervention Level

Abstract Interrupted time-series (ITS) designs are a robust and increasingly popular non-randomized study design for strong causal inference in the evaluation of public health interventions. One of the most common techniques for model parameterization in the analysis of ITS designs is segmented regression, which uses a series of indicators and linear terms to represent the level and trend of the time-series before and after an intervention. In this article, we highlight an important error often presented in tutorials and published peer-reviewed papers using segmented regression parameterization for the analyses of ITS designs. We show that researchers cannot simply use the product between their calendar time variable and the indicator variable indicating pre- versus post-intervention time periods to represent the post-intervention linear segment. If researchers use this often-presented parameterization, they will get an erroneous result for the level change in their time-series. We show that researchers must take care to use the product between their intervention variable and the time elapsed since the start of the intervention, rather than the time since the beginning of their study. Thus, the second linear segment of the time-series indexing the post-intervention level and trend should be zero before intervention implementation and begin by counting from zero, rather than counting from the time elapsed since the beginning of the study. We hope that this article can clarify segmented regression parameterization for the analysis of ITS designs and help researchers avoid confusing and erroneous results in the level changes of their time-series.

scholarly journals On ignoring instead of chasing after coagulation factor VII during warfarin management; an interrupted time series study

Blood ◽  
2021 ◽  
Author(s):  
Alma R. Oskarsdottir ◽  
Brynja R. Gudmundsdottir ◽  
Hulda M. Jensdottir ◽  
Bjorn Flygenring ◽  
Ragnar Palsson ◽  
...  
Keyword(s):  
Time Series ◽  
Coagulation Factor ◽  
Maintenance Phase ◽  
Interrupted Time Series ◽  
Factor Vii ◽  
Segmented Regression ◽  
Post Intervention ◽  
Coagulation Factor Vii ◽  
Monitoring Test ◽  
Time In Range

During warfarin management, prothrombin time (PT) based PT-INR variability is partly due to clinically inconsequential fluctuations of factor (F) VII. The new Fiix-PT and Fiix-normalized ratio (Fiix-NR), unlike PT-INR, is only affected by reduced FII and FX. Starting July 1st 2016 we replaced PT-INR monitoring of warfarin with Fiix-NR in our patients. Using interrupted time series methods, we retrospectively assessed if this affected thromboembolism (TE) and major bleeding (MB) incidence during 12 months prior to and 18 months after the replacement, months 13-18 being predefined as transitional months. The dynamic cohort comprised all our service´s 2,667 maintenance phase warfarin patients managed at any time during the 30 months. Using two-segmented regression, a breakpoint in total TE monthly incidence became evident six months after laboratory monitoring test replacement, followed by 56% reduction in TE incidence (from 2.82% to 1.23% per patient year, P=0.019 by ANOVA). Three-segmented regression found no significant TE incidence trend (slope +0.03) prior to test replacement but during months 13-18 and 19-30 the TE incidence gradually decreased (slope -0.12; R2=0.20;P=0.007). Based on segmented regressions, MB incidence (2.79% ppy) did not differ pre- or post-intervention. Incidence comparison during the 12 month Fiix- and PT-periods confirmed a statistically significant 55-62% reduction in TE. Fiix-monitoring reduced testing, dose adjustments and normalized ratio variability, and prolonged testing intervals and time in range. We conclude that ignoring FVII during Fiix-NR monitoring in real world practice stabilizes the anticoagulant effect of warfarin and associates with major reduction in thromboembolism without increasing bleeding.

scholarly journals LB-16. Association Between Universal Face Shield in a Quaternary Care Center and Reduction of SARS-COV2 Infections Among Healthcare Personnel and Hospitalized Patients

2020 ◽  
Vol 7 (Supplement_1) ◽  
pp. S851-S851
Author(s):  
Vagesh Hemmige ◽  
Becky Winterer ◽  
Todd Lasco ◽  
Bradley Lembcke
Keyword(s):  
Time Series ◽  
Care Center ◽  
Interrupted Time Series ◽  
Hospitalized Patients ◽  
Healthcare Personnel ◽  
Segmented Regression ◽  
Intervention Period ◽  
Interrupted Time Series Analysis ◽  
Post Intervention ◽  
The Impact

Abstract Background SARS-COV2 transmission to healthcare personnel (HCP) and hospitalized patients is a significant challenge. Our hospital is a quaternary healthcare system with more than 500 beds and 8,000 HCP. Between April 1 and April 17, 2020, we instituted several infection prevention strategies to limit transmission of SARS-COV2 including universal masking of HCP and patients, surveillance testing every two weeks for high-risk HCP and every week for cluster units, and surveillance testing for all patients on admission and prior to invasive procedures. On July 6, 2020, we implemented universal face shield for all healthcare personnel upon entry to facility. The aim of this study is to assess the impact of face shield policy on SARS-COV2 infection among HCP and hospitalized patients. Figure 1- Interrupted time series Methods The preintervention period (April 17, 2020-July 5, 2020) included implementation of universal face masks and surveillance testing of HCP and patients. The intervention period (July 6, 2020-July 26, 2020) included the addition of face shield to all HCP (for patient encounters and staff-to-staff encounters). We used interrupted time series analysis with segmented regression to examine the effect of our intervention on the difference in proportion of HCP positive for SARS-COV2 (using logistic regression) and HAI (using Poisson regression). We defined significance as p values < 0.05. Results Of 4731 HCP tested, 192 tested positive for SARS-COV2 (4.1%). In the preintervention period, the weekly positivity rate among HCP increased from 0% to 12.9%. During the intervention period, the weekly positivity rate among HCP decreased to 2.3%, with segmented regression showing a change in predicted proportion positive in week 13 (18.0% to 3.7%, p< 0.001) and change in the post-intervention slope on the log odds scale (p< 0.001). A total of 14 HAI cases were identified. In the preintervention period, HAI cases increased from 0 to 5. During the intervention period, HAI cases decreased to 0. There was a change between pre-intervention and post-intervention slope on the log scale was significant (p< 0.01). Conclusion Our study showed that the universal use of face shield was associated with significant reduction in SARS-COV2 infection among HCP and hospitalized patients. Disclosures All Authors: No reported disclosures

The Impact of a National Antimicrobial Stewardship Program on Antibiotic Prescribing in Primary Care: An Interrupted Time Series Analysis

2018 ◽  
Vol 69 (2) ◽  
pp. 227-232 ◽  
Author(s):  
Violeta Balinskaite ◽  
Alan P Johnson ◽  
Alison Holmes ◽  
Paul Aylin
Keyword(s):  
Primary Care ◽  
Time Series ◽  
Broad Spectrum ◽  
Interrupted Time Series ◽  
Series Data ◽  
Antibiotic Prescribing ◽  
Segmented Regression ◽  
Broad Spectrum Antibiotic ◽  
Stewardship Program ◽  
The Impact

Abstract Background The Quality Premium was introduced in 2015 to financially reward local commissioners of healthcare in England for targeted reductions in antibiotic prescribing in primary care. Methods We used a national antibiotic prescribing dataset from April 2013 until February 2017 to examine the number of antibiotic items prescribed, the total number of antibiotic items prescribed per STAR-PU (specific therapeutic group age/sex-related prescribing units), the number of broad-spectrum antibiotic items prescribed, and broad-spectrum antibiotic items prescribed, expressed as a percentage of the total number of antibiotic items. To evaluate the impact of the Quality Premium on antibiotic prescribing, we used a segmented regression analysis of interrupted time series data. Results During the study period, over 140 million antibiotic items were prescribed in primary care. Following the introduction of the Quality Premium, antibiotic items prescribed decreased by 8.2%, representing 5933563 fewer antibiotic items prescribed during the 23 post-intervention months, as compared with the expected numbers based on the trend in the pre-intervention period. After adjusting for the age and sex distribution in the population, the segmented regression model also showed a significant relative decrease in antibiotic items prescribed per STAR-PU. A similar effect was found for broad-spectrum antibiotics (comprising 10.1% of total antibiotic prescribing), with an 18.9% reduction in prescribing. Conclusions This study shows that the introduction of financial incentives for local commissioners of healthcare to improve the quality of prescribing was associated with a significant reduction in both total and broad-spectrum antibiotic prescribing in primary care in England.

The Impact of the Change in Payment System for German Psychiatric Clinics on Treatment Success: An Interrupted Time Series Analysis

2019 ◽  
Vol 82 (06) ◽  
pp. 559-567
Author(s):  
Christina Niedermeier ◽  
Andrea Barrera ◽  
Eva Esteban ◽  
Ivana Ivandic ◽  
Carla Sabariego
Keyword(s):  
Time Series ◽  
Negative Impact ◽  
Treatment Success ◽  
Interrupted Time Series ◽  
Sensitivity Analyses ◽  
Interrupted Time Series Analysis ◽  
Post Intervention ◽  
Intervention Point ◽  
The Impact ◽  
New System

Abstract Background In Germany a new reimbursement system for psychiatric clinics was proposed in 2009 based on the § 17d KHG Psych-Entgeltsystem. The system can be voluntary implemented by clinics since 2013 but therapists are frequently afraid it might affect treatment negatively. Objectives To evaluate whether the new system has a negative impact on treatment success by analysing routinely collected data in a Bavarian clinic. Material and methods Aggregated data of 1760 patients treated in the years 2007–2016 was analysed with segmented regression analysis of interrupted time series to assess the effects of the system on treatment success, operationalized with three outcome variables. A negative change in level after a lag period was hypothesized. The robustness of results was tested by sensitivity analyses. Results The percentage of patients with treatment success tends to increase after the new system but no significant change in level was observed. The sensitivity analyses corroborate results for 2 outcomes but when the intervention point was shifted, the positive change in level for the third outcome became significant. Conclusions Our initial hypothesis is not supported. However, the sensitivity analyses disclosed uncertainties and our study has limitations, such as a short observation time post intervention. Results are not generalizable as data of a single clinic was analysed. Nevertheless, we show the importance of collecting and analysing routine data to assess the impact of policy changes on patient outcomes.

scholarly journals Evaluating the impact of a hospital scale-up phase of a quality improvement intervention in Ghana on mortality for children under five

2019 ◽  
Author(s):  
Kavita Singh ◽  
Ilene Speizer ◽  
Pierre M Barker ◽  
Josephine Nana Afrakoma Agyeman-Duah ◽  
Justina Agula ◽  
...  
Keyword(s):  
Time Series ◽  
Quality Improvement ◽  
Infant Mortality ◽  
Scale Up ◽  
Interrupted Time Series ◽  
Post Intervention ◽  
Under Five ◽  
Facility Level ◽  
Provision Of Care ◽  
Southern Ghana

Abstract Objective To evaluate the scale-up phase of a national quality improvement initiative across hospitals in Southern Ghana. Design This evaluation used a comparison of pre- and post-intervention means to assess changes in outcomes over time. Multivariable interrupted time series analyses were performed to determine whether change categories (interventions) tested were associated with improvements in the outcomes. Setting Hospitals in Southern Ghana Participants The data sources were monthly outcome data from intervention hospitals along with program records. Intervention The project used a quality improvement approach whereby process failures were identified by health staff and process changes were implemented in hospitals and their corresponding communities. The three change categories were: timely care-seeking, prompt provision of care and adherence to protocols. Main outcome measures Facility-level neonatal mortality, facility-level postneonatal infant mortality and facility-level postneonatal under-five mortality. Results There were significant improvements for two outcomes from the pre-intervention to the post-intervention phase. Postneonatal infant mortality dropped from 44.3 to 21.1 postneonatal infant deaths per 1000 admissions, while postneonatal under-five mortality fell from 23.1 to 11.8 postneonatal under-five deaths per 1000 admissions. The multivariable interrupted time series analysis indicated that over the long-term the prompt provision of care change category was significantly associated with reduced postneonatal under five mortality (β = −0.0024, 95% CI −0.0051, 0.0003, P < 0.10). Conclusions The reduced postneonatal under-five mortality achieved in this project gives support to the promotion of quality improvement as a means to achieve health impacts at scale.

scholarly journals Effects of the implementation of a hand hygiene education program among ICU professionals: an interrupted time-series analysis

2019 ◽  
Vol 45 (5) ◽  
Author(s):  
Diana Marcela Prieto Romero ◽  
Maycon Moura Reboredo ◽  
Edimar Pedrosa Gomes ◽  
Cristina Martins Coelho ◽  
Maria Aparecida Stroppa de Paula ◽  
...  
Keyword(s):  
Time Series ◽  
Hand Hygiene ◽  
Education Program ◽  
Hand Hygiene Compliance ◽  
Interrupted Time Series ◽  
Interrupted Time Series Analysis ◽  
Post Intervention ◽  
Intervention Phase ◽  
Hygiene Compliance ◽  
Hygiene Education

ABSTRACT Objective: To evaluate the effects that a hand hygiene education program has on the compliance of health professionals in an ICU. Methods: This was a quasi-experimental study with an interrupted time-series design, conducted over a 12-month period: the 5 months preceding the implementation of a hand hygiene education program (baseline period); the 2 months of the intensive (intervention) phase of the program; and the first 5 months thereafter (post-intervention phase). Hand hygiene compliance was monitored by one of the researchers, unbeknownst to the ICU team. The primary outcome measure was the variation in the rate of hand hygiene compliance. We also evaluated the duration of mechanical ventilation (MV), as well as the incidence of ventilator-associated pneumonia (VAP) at 28 days and 60 days, together with mortality at 28 days and 60 days. Results: On the basis of 959 observations, we found a significant increase in hand hygiene compliance rates-from 31.5% at baseline to 65.8% during the intervention phase and 83.8% during the post-intervention phase, corresponding to prevalence ratios of 2.09 and 2.66, respectively, in comparison with the baseline rate (p < 0.001). Despite that improvement, there were no significant changes in duration of MV, VAP incidence (at 28 or 60 days), or mortality (at 28 or 60 days). Conclusions: Our findings indicate that a hand hygiene education program can increase hand hygiene compliance among ICU professionals, although it appears to have no impact on VAP incidence, duration of MV, or mortality.

Sign in / Sign up

Export Citation Format

Share Document