Increasing Contraception Documentation in an Academic Rheumatology Clinic: A Quality Improvement Approach

Abstract

Objective

This quality improvement initiative aimed to increase the rate of provider screening and documentation of contraception use for reproductive‐aged women seen in an academic rheumatology fellows’ clinic to >50% by 24 weeks, with sustained improvement at one year.

Methods

With a multidisciplinary team, we devised and implemented six interventional cycles over 24 weeks informed by key stakeholder survey responses. The primary outcome measure was the percentage of eligible visits with contraception information documented in the structured electronic health record field. The smoking status documentation rate was tracked as a balancing measure, and the percentage of contraception documenters who were non–medical doctor (MD) clinical staff, the target group for the intervention, was tracked as a process measure. Outcome, process, and balancing measures were assessed every two weeks over one year.

Results

Over 24 weeks, the rate of contraception documentation increased from 11% to 54% (median 30%), and the median smoking status documentation rate was 88% (68%–97% range); the median rate of non‐MD clinical staff documenters was 92% (70%–100% range). Interventions included an introductory educational session with documentation instruction and scripts to guide screening, email reminders from nursing leadership, and interim educational sessions. At one year, the contraception documentation rate was sustained at 50%.

Conclusion

A multicycle educational intervention led to an increase in the contraception documentation rate from 11% to 54% for reproductive‐aged women seen in an academic rheumatology fellows’ clinic over 24 weeks, with sustained improvement at one year. Future interventions will focus on increasing contraception counseling and referrals for patients in rheumatology clinics who were high risk.

INTRODUCTION

SIGNIFICANCE & INNOVATIONS.

• System‐based interventions are needed to increase contraception screening and documentation for individuals with childbearing potential seen by rheumatologists to decrease their risk of preventable adverse pregnancy outcomes and teratogenicity.

• Our quality improvement initiative incorporated stakeholder feedback and standardized, automated processes to increase contraception screening and documentation for all reproductive‐aged female patients in an academic rheumatology clinic.

• We increased contraception documentation for patients who were at risk from 11% to 54% over a 24‐week primary intervention period and demonstrated sustained improvement over the course of one year.

Systemic rheumatic diseases (SRDs) often affect reproductive‐aged women and confer risk of various adverse pregnancy outcomes (APOs), especially when disease is active early in pregnancy. ¹ , ² , ³ , ⁴ In addition, teratogenic medications are frequently used to treat SRDs. The use of effective contraception can prevent medically ill‐timed pregnancies, mitigating the risk of APOs and teratogenicity. Guidelines recommend that rheumatologists counsel reproductive‐aged women with SRDs about contraception use and family planning. ⁵ , ⁶ Published quality indicators for systemic lupus erythematosus (SLE) management include documentation of contraception counseling when prescribing teratogenic medication. ⁷

Studies report mixed findings regarding the consistency with which patients who are at risk receive counseling. In a single‐center retrospective study including 478 reproductive‐aged women with SLE, 375 (78.5%) received contraception counseling from at least one specialty and 272 (56.9%) from rheumatology. ⁸ Women who were prescribed teratogenic medications had approximately twice the odds of receiving contraception counseling from at least one subspecialty (odds ratio [OR] 2.01, 95% confidence interval [CI] 1.23–3.26) and from rheumatology (OR 1.86, 95% CI 1.27–2.73) compared to those who were not. ⁸ However, in another survey study including 86 reproductive‐aged women with SLE at risk of unplanned pregnancy, only 41% had received contraception counseling in the preceding year, and only 37% of those were prescribed teratogenic medications; there were no significant differences regarding receipt of contraception counseling, frequency of reported contraception use, or reported type of contraception used reception between those who were prescribed teratogenic medications and those who were not. ⁹

Women with SRDs are often not prescribed effective contraception. In a 2019 single‐center study including 2,455 reproductive‐aged women with SRD, 72% were prescribed a known teratogen during the two‐year study timeframe, and only 32% were prescribed contraception; there was no difference in contraception prescriptions between those who were prescribed a teratogenic medication and those who were not. ¹⁰ In a Medicaid claims‐based study of 24,693 reproductive‐aged women with SLE from 2000 to 2010, only 10% received any contraception and 2% received highly effective contraception, with no difference based on teratogenic medication reception. ¹¹

Documentation regarding reported contraception use is inadequate and poorly standardized. A 2021 national electronic health record (EHR)‐based study using data from the Rheumatology Informatics Systemic for Effectiveness (RISE) registry reported that <10% of 28,835 reproductive‐aged women with SLE or rheumatoid arthritis had contraception information documented in a structured data field, and rates were low across the 212 included practices. ¹² Consistent with national data, in the Hospital for Special Surgery (HSS) adult rheumatology fellows’ clinic, only 11% of female patients aged 18 to 45 years had contraception information documented in the structured EHR field between July 1, 2020, and June 30, 2022. Inadequate documentation undermines efforts to systematically measure and improve upon this gap in health care delivery. To address this issue locally, we launched a quality improvement (QI) initiative with a multidisciplinary team including key stakeholders: rheumatology attending and fellow physicians, nursing staff, nursing leadership, and a designated nursing champion. The goal was to increase contraception screening and documentation in the HSS adult rheumatology fellows’ clinic to 50% by 24 weeks and to demonstrate sustained improvement over one year.

MATERIALS AND METHODS

We administered an anonymous survey to all adult rheumatology fellows and nursing staff before the intervention, which included demographics (ie, age, gender, race and ethnicity), professional role, and perceived barriers and facilitators to documenting contraception information in the EHRs. Responses to the stakeholder survey question, “Please select any barriers you have encountered or reasons for not entering patient information into the sexual activity field in the EHR (Choose all that apply),” were tabulated in a Pareto chart; responses to the question, “Which of the following would help you to ask patients about contraception use and document this information in the sexual activity field in the EHR (Choose all that apply),” provided insight into facilitators. Due to the small number of respondents and efforts to maintain anonymity, demographic characteristics including professional role were separated from responses to questions about barriers and facilitators to documentation before analysis. This information informed the development of iterative Plan–Do–Study–Act (PDSA) cycles.

Clinical staff were instructed to document contraceptive method by selecting from a comprehensive list in an existing structured EHR field, including permanent methods and reversible methods with varying degrees of effectiveness; more than one option could be selected, and additional information could be written in an open field if indicated. The primary outcome measure was the percentage of female patients aged 18 to 45 years seen in the HSS adult rheumatology fellows’ clinic with any contraception information documented. The overall aim of the initiative was to increase the rate of provider documentation of contraception information in the structured EHR field for female patients aged 18 to 45 years seen in the HSS adult rheumatology fellows’ clinic from 11% (the mean frequency of documentation in the preceding two years, July 1, 2020, to June 30, 2022) to 50% over 24 weeks. Secondarily, we aimed to demonstrate sustained improvement from the start of the initiative through one year (52 weeks).

The intention was to incorporate contraception screening and documentation into the rooming workflow, in which a member of the nursing staff rooms the patient and reviews and documents elements of the patient's medical and social history before the physician's examination. We assessed the percentage of those who documented the information who were nursing staff (ie, had professional degrees of registered nurse [RN], patient care assistant [PCA], or licensed practical nurse) as a process measure to confirm our interventions reached the intended group. Our balancing measure was the percentage of the target population with smoking status documented in the corresponding structured EHR field. We chose smoking status documentation as a balancing measure because it was an established part of the rooming workflow and was routinely documented by nursing staff before launching this QI initiative. We monitored this measurement to ensure that adding a step to the workflow would not detract from documentation of another key metric.

We conducted six PDSA cycles over the 24‐week primary intervention period, including focused educational sessions with clinic staff and rheumatology fellows and progress updates presented during staff meetings and sent via email (Table 1). During educational sessions for staff, we highlighted the fact that contraception status should be asked along with other elements of a routine social history, such as smoking status, documentation of which in the same EHR section is already established practice. We provided suggested scripts when asking patients about contraception use, including the following normalizing statement: “We ask all patients seen in the rheumatology clinic nonmedical questions that may be relevant to their health, which include routine questions about sexual activity and contraception.”

Table 1.

PDSA cycles targeting barriers and incorporating facilitators identified by stakeholders over 52‐week period^*

PSDA cycle	Time point	Date	Interventions
24‐week primary intervention period
1	Week 0	12/8/23	Educational session for clinic staff Educational session for adult rheumatology fellows Stakeholder survey administered
2	Week 6	1/18/23	Reminder email from nursing leadership to clinic staff
3	Week 10	2/16/23	Progress update and educational session at clinic staff meeting
4	Week 11	2/22/23	Progress update and educational session for adult rheumatology fellows Division of Rheumatology Grand Rounds focused on importance of contraception and considerations for pregnancy termination in patients with rheumatic disease
5	Week 19	4/24/23	Reminder email with progress update to clinic staff
6	Week 22	5/18/23	Progress update and educational session at clinic staff meeting
28‐week secondary intervention period/monitoring for sustainability
7	Week 37	8/25/23	Introduction/reminder to existing and new adult rheumatology fellows
8	Week 43	10/6/23	Reminder email to clinic staff

^*PDSA, Plan–Do–Study–Act.

We tabulated outcome, balancing, and process measures every two weeks for a six‐week run‐in period to measure baseline information and optimize our data collection processes. We continued recording these measures every 2 weeks over the subsequent 24 weeks. Following the primary intervention period, we delivered an introductory educational session to nursing and physician staff at the start of a new academic year followed by an email reminder six weeks later and continued to monitor outcome, balancing, and process measures every 2 weeks through the 52‐week endpoint without further intervention.

We descriptively analyzed aggregate patient‐ and visit‐level data over the one‐year timeframe. Using the chi‐square test, we compared the frequency of contraception documentation between those who had and had not been prescribed a teratogenic medication at any point during the study period. Stata version 18 was used to calculate descriptive statistics and a P value <0.05 was considered statistically significant. This study was approved by the HSS Institutional Review Board (2022‐2112).

RESULTS

Of 19 clinic nursing staff and adult rheumatology fellows, 17 (89%) responded to the stakeholder survey: 47% rheumatology fellows, 35% RNs, and 18% PCAs; 65% female; 47% White, 29% Asian/Indian, 6% Black, and 12% other race; and 82% not Hispanic/Latino. The most important barriers to contraception documentation identified—accounting for 80% of survey responses (the “vital few” in the Pareto chart; Figure 1)—included lack of direction (ie, “I have not been told I should ask patients these questions.”) and awareness (ie, “I do not know how to find this field in the patient chart,” and “I didn't know that there was a sexual activity field until now.”). The remaining 20% of responses (the “trivial many” in the Pareto chart) included time, comfort level discussing contraception with patients, and disagreement with the premise that these questions should be routinely asked in a rheumatology clinic (Figure 1). Facilitators provided by stakeholders included education about family planning and contraception considerations for patients in rheumatology clinics (n = 8), dedicated time (n = 7), scripts to guide the conversation with patients (n = 5), technical instructions on navigating to and entering information in the EHR field (n = 4), and being instructed to screen and document this information (n = 3). Educational content provided over the intervention period addressed these variables.

Figure 1.

Pareto chart indicating barriers to contraception documentation elicited from stakeholder survey.

We tracked and analyzed primary outcome and balancing measures in a run chart (Figure 2). The primary outcome measure—the percentage of eligible patients with contraception documentation in the EHR field—was 11%, 9%, and 25% at run‐in points 1, 2, and 3, respectively. The first PDSA cycle began at run‐in point 3 (week 0). Following introductory educational sessions at week 0 and a reminder email sent from nursing leadership at week 6, the contraception documentation rate reached 40% by week 10. A progress update and additional educational sessions were followed by a downtrend in the contraception documentation rate to 22% by week 16. This downtrend over the course of four weeks began during winter holidays. Additional reminders and progress updates via email and at a clinical staff meeting at weeks 19 and 22 were followed by an increase in the contraception documentation rate to a peak of 54% by week 24. The median contraception documentation rate over the initial 24‐week intervention period was 30%. The primary balancing measure—the percentage of patients with smoking status documented in the structured field—was 81%, 86%, and 89% at run‐in points 1, 2, and 3, respectively, and 90% at 24 weeks (median 87%, range 68%–97%). The percentage of documenters who were nursing staff was ≥70% (median 92%, range 70%–100%) throughout the initial 24‐week intervention period; the remainder were rheumatology fellows.

Figure 2.

Run chart tracking primary outcome measure (contraception documentation rate) and balancing measure (smoking status documentation rate) every 2 weeks over a 52‐week period. PDSA, Plan–Do–Study–Act. Red dashed line: 80% cut‐off; orange solid line: cumulative percentage.

During the following 28‐week period, the increased contraception documentation rate was sustained. Between weeks 24 and 34, the rate declined from 54% to a 33% nadir. This downtrend coincided with the entrance of a new class of fellows. Additional reminders and updates to the clinic staff at weeks 37 and 43 led to a peak of 65% documentation at week 44. By week 52, the contraception documentation rate was 50%. The median contraception documentation rate over the final 28‐week intervention week period was 49%. The rate of smoking status documentation was 100% at week 52 (median 93%, range 84%–100%).

There were 402 female patients aged 18 to 45 years seen in the adult rheumatology fellows’ clinic for 897 total visits between December 9, 2022, and December 7, 2023. The mean ± SD age was 33.87 ± 7.35 years. In aggregate, 193 of 402 patients (48.01%) had any contraception information documented, and 147 of 402 patients (36.57%) had an active prescription for at least one of the following teratogenic medications (not mutually exclusive): leflunomide (n = 8), lenalidomide (n = 1), methotrexate (n = 58), mycophenolate (n = 61), warfarin (n = 8), or angiotensin‐converting enzyme inhibitor/angiotensin receptor blocker (n = 47). There was no significant difference in the frequency of contraception documentation between individuals prescribed versus not prescribed a teratogenic medication (73 of 147 [49.66%] vs 120 of 255 [47.06%], P = 0.62).

DISCUSSION

With implementation of a multicycle intervention targeting barriers reported by clinic staff, we achieved our primary aim: standardized contraception documentation increased from 11% to 54% by week 24, with sustained improvement at one year. Interventions included educational sessions and progress updates that included specific instruction to nursing staff and physicians about how, when, and where to document sexual activity and contraception information, education regarding family planning and contraception considerations for patients in the rheumatology clinic, and scripts to guide conversations with patients. We demonstrated that the smoking status documentation rate did not decline, from 81% at the outset to 90% at the final time point, suggesting that this added step was not overly burdensome or detracting from other important aspects of the rooming process. Because the majority (≥70%) of documenters at each time point were nursing staff, we demonstrated that the intervention reached the intended group and step in the clinic workflow. Of 402 eligible patients during the 52‐week period, approximately half had any contraception information documented in the structured field. Approximately one‐third had an active prescription for at least one teratogenic medication. Documentation rates were no different between those who were prescribed a teratogenic medication and those who were not, suggesting that clinicians were routinely screening and documenting contraception information regardless of risk factors.

Previous QI initiatives in academic rheumatology practices have aimed to improve contraception documentation, counseling, and related provider actions for reproductive‐aged women who were prescribed teratogenic medications, with varying success. ¹³ , ¹⁴ , ¹⁵ Sadun et al ¹³ conducted a 10‐month, multicycle intervention in an academic rheumatology clinic using provider education and reminders, modification of EHR templates, and patient questionnaires to improve contraception counseling and subsequent provider action for female patients aged 18 to 45 years who were prescribed a teratogenic medication. The investigators demonstrated increased contraception documentation from a higher baseline of 46% to 64%, increased contraception counseling from 30% to 45%, and increased provider action from 33% to 46%. Wolfgang et al ¹⁵ aimed to increase documentation of pregnancy and contraception counseling by rheumatology providers in an academic clinic for female patients aged 17 to 50 years who were prescribed teratogenic medications. Following a six‐month intervention period—which involved an informational session, administering provider education handouts, posting signs in examination rooms, and sending reminder emails—the percentage of eligible patients with contraception counseling documented anywhere in notes or via billing codes increased from 37% to 51%. Our intervention targeted a broader patient population—all female patients aged 18 to 45 years regardless of teratogenic medication prescription—because there are many other clinical scenarios in which family planning and avoiding unplanned pregnancy may be important for patients in rheumatology clinics with reproductive potential, especially for those with active SRDs.

In a six‐month QI initiative aimed at increasing standardized pregnancy intention screening by encouraging rheumatologists to ask reproductive‐aged women One Key Question (OKQ; “Would you like to become pregnant in the next year?”), a trademarked clinical tool developed for primary care providers, Pryor et al ¹⁴ reported low uptake: only 83 of 957 eligible patients (9%) had OKQ documented in their EHR after the intervention. Patients with any documentation of pregnancy intention screening, including with OKQ, were more likely than those who were not screened to have contraception documented and to be referred to obstetrics/gynecology for follow‐up. ¹⁴ As opposed to these projects that relied on chart review for measurement of various outcomes, ¹³ , ¹⁴ , ¹⁵ we focused on systematizing and normalizing contraception screening for all eligible patients regardless of pregnancy intent and standardizing documentation in a structured field, enabling automated measurement.

There were several strengths of our intervention. Specific PDSA cycles were devised in response to a baseline survey of key stakeholders, including the group primarily tasked with screening and documentation. Our initiative involved iterative educational sessions, updates, and reminders during the 24‐week primary intervention period, leading to substantial improvement in contraception documentation, followed by a 28‐week period of minimal intervention, during which time we demonstrated a sustained shift in practices. With frequent, automated measurement of balancing and process measures, we confirmed that the interventions impacted the behavior of the target group and did not lead to unintended negative consequences. Priorities of this initiative were to incorporate the intervention into an existing workflow and to avoid the need for manual chart review in order to ensure both improvement and long‐term sustainability.

There were several limitations of our intervention. The target patient was defined by age of 18 to 45 years and female sex, as determined by how patients are registered in our EHRs, which may not consistently reflect reproductive potential. Following the baseline stakeholder survey, we did not have a formal process for obtaining and addressing ongoing feedback. Given the anonymous nature of the survey, we were also limited in our ability to address specific barriers based on professional role. In addition, we did not solicit patient input. Lastly, although improvement was sustained, we did not see a continued upward trend in contraception documentation during the secondary intervention period, suggesting that there are important, persistent barriers our intervention did not address.

Studies have identified several barriers to appropriate contraception counseling by rheumatologists for patients with SRDs, including limited knowledge, confidence, time, and questions about their role relative to primary care providers and gynecologists. ¹⁶ , ¹⁷ Because patients with SRDs face disease‐ and treatment‐related risks in pregnancy, it is important for rheumatologists to take ownership over discussions about contraception for pregnancy prevention and for timing pregnancies such that risks are mitigated. Beyond guideline recommendations to this effect, ⁵ , ⁶ , ⁷ patients expect their rheumatologists to discuss sexual and reproductive health topics with them and to proactively collaborate with their gynecologists. ¹⁸

The focus of this QI initiative was on contraception screening and documentation by rheumatology clinicians (primarily nursing staff), not on contraception counseling. However, clinical staff elicited and documented the specific contraceptive method used by patients; this information is critical for individualizing rheumatologists’ counseling and recommendations for patients because the effectiveness of different methods varies considerably, and there are medical contraindications to certain methods for select patient groups. Our interventions led to a notable increase in the percentage of eligible visits with any form of contraception documented and sustainably improved the clinical workflow. Establishing this standardized practice of contraception screening and documentation will enable future research and QI studies focused on improving rheumatologists’ counseling and referrals to women's health providers for patients in need of prescription contraception with the goal of optimizing patients’ contraception practices and related health outcomes.

AUTHOR CONTRIBUTIONS

All authors contributed to at least one of the following manuscript preparation roles: conceptualization AND/OR methodology, software, investigation, formal analysis, data curation, visualization, and validation AND drafting or reviewing/editing the final draft. As corresponding author, Dr Siegel confirms that all authors have provided the final approval of the version to be published, and takes responsibility for the affirmations regarding article submission (eg, not under consideration by another journal), the integrity of the data presented, and the statements regarding compliance with institutional review board/Declaration of Helsinki requirements.

Supporting information

Disclosure Form: