The Effect of an EHR Order Set on Cancer Screening Order Rates in Community-Based Health Centers

Abstract

Adoption of electronic health record (EHR)-based clinical decision support tools in community-based health centers might increase the provision of indicated cancer screening orders. We examined: (1) if the use of the care gaps smartset (CGS), an EHR tool that expedites ordering care, is associated with colorectal/cervical cancer (CRC/CVC) screening order rates; and (2) how selected implementation strategies, barriers, and facilitators impact CGS use.Within a sequential mixed methods design, we used multivariate regression to assess associations between clinic- and provider-level CGS use and cancer screening order rates. Tool use rates (3/2018-12/2023) were measured as the rate of encounters at which any orders were placed via the CGS and then categorized by use level. Surveys (n = 81) and semi-structured interviews (n = 11) with clinic staff assessed strategies to improve tool use.Clinics and providers that ever used the CGS had higher CRC screening order rates than non-users. Higher CGS use was associated with better CRC screening order rates. By 12/2023, CRC screening orders were 4.4% (p < 0.05) higher in high-use clinics versus those with no CGS use. CGS use was not associated with CVC screening order rates. Qualitative findings indicate effective CGS use was enhanced by leadership support, clear workflows, and clinic-led training. Barriers to CGS use included low user awareness of/trust in the tool, and tool functions that were not optimized.CGS use can support cancer screening ordering; its adoption may be enhanced by varied training approaches and workflow design.

Fig. 1

Summary of care gap smartset use barriers (X) and facilitators (✓) and opportunities for improvement.

Fig. 2

Clinic-level CGS use and colorectal cancer screening order rates, 2018–2023. Implementation year is defined by the year start date as follows; 0 = 03/2018–02/2019 with 458 clinics (before the CGS tool became available), 1 = 03/2019–02/2020 with 608 clinics, 2 = 03/2020–07/2020 with 655 clinics (year 2.5 = 08/2020–02/2021, 655 clinics), 3 = 03/2021–02/2022 with 818 clinics, and 4 = 03/2022–12/2023 with 939 clinics. Mean CRC order rates for the Implementation year (% patients with orders placed within 30 days of an encounter when due for CRC screening) were modeled from a Poisson regression of monthly rates adjusting for clinics in their first year on EHR and accounting for within clinic correlation using robust sandwich estimators. CGS, care gaps smartest; CRC, colorectal; EHR, electronic health record.

Fig. 3

Clinic-level CGS use and cervical cancer screening order rates for women (age <30 years), 2018–2023. Implementation year is defined by the year start date as follows; 0 = 03/2018–02/2019 with 427 clinics (before the CGS tool became available), 1 = 03/2019–02/2020 with 553 clinics, 2 = 03/2020–07/2020 with 584 clinics (year 2.5 = 08/2020–02/2021, 601 clinics), 3 = 03/2021–02/2022 with 760 clinics, and 4 = 03/2022–12/2023 with 890 clinics. Mean CVC order rates for the implementation year (% of patients with order placed within 30 days of an encounter when due for cervical cancer screening) were modeled from Poisson regression of monthly rates adjusting for providers in their first year on EHR and accounting for within provider correlation using robust sandwich estimators. CGS, care gaps smartest; CVC, cervical cancer; EHR, electronic health record.

Fig. 4

Clinic-level CGS use and cervical cancer order rates for women (age ≥30 years), 2018–2023. Implementation year is defined by year start date as follows; 0 = 03/2018 with 465 clinics (before the CGS tool became available), 1 = 03/2019 with 614 clinics, 2 = 03/2020 with 655 clinics, 3 = 03/2021 with 818 clinics, and 4 = 03/2022 with 967 clinics. Mean PAP order rates for the implementation year (% of patients with order placed within 30 days of an encounter when due for cervical cancer screening) were modeled from Poisson regression of monthly rates adjusting for providers in their first year on EHR and accounting for within provider correlation using robust sandwich estimators. CGS, care gaps smartest; EHR, electronic health record.