PUBH 8503 Week 11 Approach Chapter Example

The Term's Decisions in Conventional Order: An Approach Chapter for a Segmented Regression Study of an Invented Clinic's Program to Bring Treated Chlamydia Clients Back for Retesting

Student Name

Walden University

PUBH 8503: Scholar-Practitioner Inquiry III

Instructor Name

Month Day, Year

The Term's Decisions in Conventional Order: An Approach Chapter for a Segmented Regression Study of an Invented Clinic's Program to Bring Treated Chlamydia Clients Back for Retesting

Design

This study uses an interrupted time series to evaluate a follow-up program at Harbor Street, the walk-in clinic followed through this inquiry sequence. The program confirms contact details at the treatment visit, texts and calls each client treated for chlamydia when a retest falls due, and offers mailed self-collection kits to clients in two outlying neighborhoods. The design compares the level and trend of a monthly outcome before and after the program began, using segmented regression to separate any change at the switch from the trend already under way (Wagner et al., 2002). A comparison with a neighboring clinic was considered and set aside because the data agreement could not be completed in time, the neighboring clinic planned its own reminder change, and the two clinics differ in size and population.

Setting

Harbor Street is a public clinic for sexual health in a city of moderate size, and most of its clients walk in. It is described here without details that would identify it. Its laboratory processes chlamydia tests on site, and its outreach staff already contact clients about test results.

Population

The study includes every client with a positive chlamydia result at the clinic during the study window. Clients known to be pregnant at the time of the result are excluded, since pregnancy carries its own testing schedule (Workowski et al., 2021). A client with a second positive result within 60 days of a first is counted once, at the first result, so that repeat tests shortly after treatment do not create new cohort members.

Data Sources

Three clinic sources are used: the laboratory information system, the visit record and the outreach log. The laboratory system and visit record link exactly by internal client number; the outreach log uses the same number and was spot-checked for typing errors. State laboratory reports are not used, because linkage was not approved within the timeline. The laboratory changed test order codes 20 months before the switch, so retests are identified by both old and new codes.

Outcome

The primary outcome is the monthly share of clients with a positive chlamydia result who have a further chlamydia result in the clinic laboratory's records two to four months later. Each client belongs to the month of the original positive result. Retests at other laboratories are not counted.

Time Points

The series runs for 36 monthly cohorts before the switch and 12 after. Data collection therefore closes 16 months after the switch, giving the last cohort its full window. The switch month is fixed by the first entry in the visit record's new contact-confirmation field, which provides an objective date.

Analysis

Segmented regression will estimate three quantities: the slope before the switch, the change in level at the switch and the change in slope after it. Monthly data will be checked for autocorrelation and for seasonal patterns, since retesting may dip in winter and around holidays; seasonal terms will be added if needed (Lopez Bernal et al., 2017). The month of the switch will be excluded in a sensitivity analysis, because it mixes clients treated before and after the program began.

A control series strengthens the design. The monthly share of clients treated for syphilis who complete follow-up blood tests is recorded in the same laboratory, is subject to the same clinic-wide events and records changes, and is untouched by the program. A change at the switch in the chlamydia series but not in the syphilis series would support a program effect; a change in both would point to a clinic-wide cause (Lopez Bernal et al., 2018).

A subgroup analysis by visit type, walk-in against scheduled, will be reported with caution, because visit type comes from the visit record, which is being replaced in the same quarter.

For the quality committee, the same monthly series will be shown as a run chart with its median, applying the usual rule that a run of at least six months on one side of the median marks a shift (Perla et al., 2011).

Ethics

The study uses records collected during care. Identifiers are removed by the clinic's data manager before the data reach the researcher. Whether the study is exempt is for the university review board to decide under the federal rules on secondary research with de-identified records (45 C.F.R. § 46.104). A data use agreement with the clinic governs storage and destruction of the extract.

Limitations

Without a concurrent comparison clinic, an outside event coinciding with the switch cannot be fully excluded; the control series and documented clinic changes reduce but do not remove this risk. Retests at other laboratories are uncounted, so the outcome measures return to Harbor Street rather than retesting in general; any client the reminders sent to a different laboratory would be missed. Months with few treated clients will yield unstable shares. The design cannot say why clients still miss retests, a question left for a later qualitative study. Results from one clinic may not apply to clinics with different populations or schedules, and the threats a single-site quasi-experiment leaves open should be stated plainly in any report (Shadish et al., 2002).

Intervention Fidelity

The outreach log records, for each treated client after the switch, whether contact details were confirmed, when the text and call were made, and whether a kit was mailed and returned. These records allow the study to report how fully the program was delivered month by month. If delivery was weak in some months, the analysis will note it, since a program that was not delivered cannot be expected to change the series. Fidelity is reported descriptively and is not used to select which months enter the main analysis, which would bias the estimate toward the months that went well.

Dissemination

Results will go first to the medical director and quality committee as a run chart and short report, then to the dissertation committee in full, and finally to the state STI program, whose linkage would allow a fuller evaluation in future.

References

Exemptions, 45 C.F.R. § 46.104 (2018).

Lopez Bernal, J., Cummins, S., & Gasparrini, A. (2017). Interrupted time series regression for the evaluation of public health interventions: A tutorial. International Journal of Epidemiology, 46(1), 348-355. https://doi.org/10.1093/ije/dyw098

Lopez Bernal, J., Cummins, S., & Gasparrini, A. (2018). The use of controls in interrupted time series studies of public health interventions. International Journal of Epidemiology, 47(6), 2082-2093. https://doi.org/10.1093/ije/dyy135

Perla, R. J., Provost, L. P., & Murray, S. K. (2011). The run chart: A simple analytical tool for learning from variation in healthcare processes. BMJ Quality & Safety, 20(1), 46-51. https://doi.org/10.1136/bmjqs.2009.037895

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Houghton Mifflin.

Wagner, A. K., Soumerai, S. B., Zhang, F., & Ross-Degnan, D. (2002). Segmented regression analysis of interrupted time series studies in medication use research. Journal of Clinical Pharmacy and Therapeutics, 27(4), 299-309. https://doi.org/10.1046/j.1365-2710.2002.00430.x

Workowski, K. A., Bachmann, L. H., Chan, P. A., Johnston, C. M., Muzny, C. A., Park, I., Reno, H., Zenilman, J. M., & Bolan, G. A. (2021). Sexually transmitted infections treatment guidelines, 2021. MMWR Recommendations and Reports, 70(4), 1-187. https://doi.org/10.15585/mmwr.rr7004a1