Quantitative and qualitative impact of One Key Question on primary care providers’ contraceptive counseling at routine preventive health visits

Abstract

Objectives:

One Key Question (OKQ) is a clinical screening tool to assess pregnancy desire in the next year. We aimed to 1) describe the effect of OKQ implementation on contraceptive counseling rates at preventive health visits and 2) evaluate primary care providers’ perception of OKQ implementation on their contraceptive counseling practices.

Study design:

We performed a quantitative retrospective chart review of preventive health visits at eight federally qualified health centers in Utah between 2014 and 2017. Implementation of OKQ included a brief training and inclusion of OKQ in the electronic medical record. Providers received OKQ training in August 2015 and re-training in March 2017. We assessed OKQ and contraceptive counseling documentation rates using interrupted-time-series analysis. We then conducted semi-structured interviews with providers and queried them about the impact of OKQ. We identified dominant themes using modified grounded theory to create an explanatory framework.

Results:

Abstracting 6634 charts yielded 9840 visits with 56 unique providers (51% physician assistant, 34% physician, 14% nurse practitioner). Interrupted-time-series analysis showed a documentation increase of OKQ in late 2015 (2.6%) and again in spring 2017 (9%), however rates remained low. Contraceptive counseling rates (39.7%) did not change after OKQ implementation. Charts with evidence of a current contraceptive method were less likely to have a OKQ response documented. Interviewees reported OKQ’s algorithm did not alter their contraceptive counseling.

Conclusions:

OKQ did not change documented rates of contraceptive counseling and uptake was low in quantitative and qualitative analyses. Our study suggests limited usefulness of OKQ in the primary care setting.

Implications:

Implementation of the One Key Question tool through training and optional EHR field did not increase documented rates of contraceptive counseling in a large federally qualified health center or affect provider contraceptive counseling. Our study suggests limited usefulness of OKQ as a robust screening tool in this primary care setting.

1. Introduction

The Centers for Disease Control and Prevention, the Department of Health and Human Services, and multiple medical academic societies recommend routine assessment of patients’ reproductive life plans and provision of patient-centered preconception and contraception care [1–3]. Despite these formal recommendations, reproductive health needs are not fully addressed during ambulatory care appointments. Only 14% of ambulatory visits with non-pregnant women include preconception or contraceptive counseling [4]. This lack of reproductive counseling disproportionately affects women from marginalized backgrounds, with unintended pregnancy (mistimed or unwanted) rates highest among people living in poverty and people of color [5]. Primary care providers in federally qualified health centers (FQHCs) constitute a vital source of contraceptive care for these populations. However, even at FQHCs, patients face multiple barriers to obtain contraceptive counseling [6]. Previous research demonstrates many provider- and patient-identified obstacles to reproductive health counseling including cost, time constraints, inadequate training, and lack of clinician and patient knowledge [7,8].

One strategy to encourage reproductive health care for all patients is implementation of One Key Question (OKQ) [9]. The Oregon Foundation for Reproductive Health developed OKQ as a primary care clinical tool to facilitate proactive screening and dialogue between patient and clinician regarding a patient’s reproductive goals. A single screening question: “Would you like to become pregnant in the next year?” is asked. Based on a response of “yes,” “no,” “ok either way,” or “unsure,” providers offer contraceptive counseling, preconception counseling, or both, respectively [10–12]. When compared to validated tools such as the Desires to Avoid Pregnancy scale, OKQ does reliably identify patients’ categorical wishes, which can be useful for initiating a conversation about contraceptive needs [12].

While the OKQ can increase contraceptive counseling rates and patient satisfaction [13,14], there is limited research on clinician attitudes towards OKQ or its effect on contraceptive counseling rates in a high-volume health system. This study aimed to 1) determine the effect of OKQ on contraceptive counseling rates during preventive health visits and 2) evaluate primary care providers’ perception of OKQ implementation on their contraceptive counseling practices.

Although simplistic by design, this simplicity limits OKQ’s ability to adequately assess the complexities of pregnancy decision-making and pushes patients towards a categorical response. As our understanding of the continuum of pregnancy decision-making and the intricacies of planning or not planning a pregnancy evolve, so too must our tools; OKQ has been criticized for lacking this nuance [15–20]. We acknowledge these imperfections of OKQ as a tool and focus our objective in this study to evaluating the uptake and use of this tool within primary care and not its quality.

2. Materials and methods

We performed a quantitative and qualitative study at eight clinics within the Community Health Centers, Inc. in Salt Lake County, UT. The Community Health Centers are FQHCs which serve approximately 3000 people with the capacity to become pregnant of reproductive age per year, approximately 75% of whom identify as Latinx. The University of Utah IRB approved all components of this study, and the Community Health Centers Board of Directors permitted recruitment of practitioners and chart access for review. This assessment of OKQ in preventive care is a sub-analysis of a larger project on the Zika virus and contraceptive counseling [21].

In August 2015, a representative from the Oregon Foundation for Reproductive Health provided a OKQ training at a Community Health Centers provider meeting. This included a 15-minute presentation on use of the tool that all current providers attended. Clinical leadership then introduced a OKQ template into the electronic health record for all preventive health visits. The electronic health record OKQ template did not provide a prompt and individual providers decided to complete it or not. Some providers delegated OKQ to their medical assistant to ask during patient rooming. A brief internal audit of charts in February 2017 showed low use of the OKQ template and clinical leadership led a second training in March 2017.

A note on gendered terms used within the methods and results: we acknowledge the diversity of gender identities in people with the capacity to become pregnant. Only charts with biologic sex selected as “female” were included in this study and we did not collect information on patient gender. Therefore, we will be using the term “female” when necessary to refer to individuals included in the chart review. Providers in the qualitative portion self-identified gender and is reported as such.

2.1. Assessment of impact of OKQ on contraceptive counseling rates

We conducted a retrospective review of preventive health visits at the Community Health Centers using their electronic health record system, eClinicalWorks. We identified potential records based on preventive care CPT codes (99385, 99395, 99384, 99394, 99386, 99396). Eligible records for review included any new or established non-pregnant female patient between ages 15 to 49 who presented for preventive care between January 1, 2014, and December 31, 2017. Exclusion criteria included history of hysterectomy or bilateral tubal ligation. Variables included the presence of documentation of OKQ, contraceptive counseling independent of OKQ, and the current contraceptive method in use. We also collected patient demographics.

To determine the effect of the introduction of OKQ on contraceptive counseling rates, we divided the study into the pre-OKQ period (P1: January 1, 2014–July 31, 2015), the initial OKQ period (P2: August 1, 2015–February 28, 2016), and the OKQ re-training period (P3: January 1, 2017–December 31, 2017). In order to assess the effect of the intervention periods on the outcome of OKQ documentation, we utilized an interrupted time series analysis approach [22]. We assigned the periods above to each patient visit based on their visit date and assigned the 6 months following the start of a period as a transition period for the interventions. We modeled the documentation outcome using a generalized linear mixed-effects model with a logit link, including the following covariates of interest: time period indicators and a linear effect for date. We used a likelihood ratio test to assess whether the date slope and date-by-period interactions improved model fit. During model selection, for the OKQ outcome only, we excluded models with a date slope since we did not expect a time trend in the OKQ documentation rate prior to the policy initiation. We did allow an intercept, due to the one documented case before the defined start period. The following covariates are also included in the models for both outcomes: documentation of any contraceptive currently in use, visit number, provider type, and provider gender. Finally, we included a random effect by patient ID and a random date slope by provider to account for correlation in outcomes between visits of the same patient, correlation of outcomes of the same provider, and the possibility of different rate of change in behavior by provider over time. We made the date variable numeric by setting the first date, 2014-1-1, to zero and then added one for each subsequent date. We then scaled the dates (subtracted mean and divided by standard deviation = Date (scaled)) to support the convergence of the generalized linear mixed-effects model (i.e., we put the date covariate on a similar scale to other covariates in the model). We conducted the interrupted time series analysis using the glmer function in from the lme4 package in R version 3.6.2 [23].

2.2. Assessment of primary care providers’ perception of OKQ implementation on their practice

We performed semi-structured, qualitative interviews with primary care providers from the Community Health Centers. Between June 4, 2018, and August 9, 2018, we recruited providers at the Community Health Centers as part of a more extensive study on Zika and contraceptive counseling [21]. Recruitment occurred through convenience sampling by emails sent to each eligible provider. Eligible participants were Family physicians, nurse practitioners, or physician assistants who provided preventive health care to females of reproductive age. A priori, we excluded three physician participants who lacked board certification in Family Medicine (two pediatricians and one obstetrician/gynecologist) from the qualitative portion, due to small numbers. A total of 40 providers were eligible for participation. Recruitment continued until we reached thematic saturation in the parent study on Zika and contraceptive counseling.

Participants took part in 20-to-30-minute, semi-structured, in-depth telephone interviews conducted by an experienced research assistant who then distributed $50 Amazon gift codes upon interview completion. We audio-recorded and professionally transcribed the de-identified interviews. We based the interview guide for the primary analysis on Zika and contraceptive counseling on the Health Belief Model [24] and piloted the interview with two obstetric fellowship family physicians who were then excluded from participation. For the OKQ sub-analysis, we asked participants open-ended questions on the impact of OKQ on their contraceptive counseling practices and utilized modified grounded theory to guide our analysis [25,26].

Initial coding occurred, using Dedoose software, on the first five interviews with three trained (JK, AT, AE) researchers engaging collaboratively in code generation (version 8.3.35 (2020), Los Angeles, CA). Two researchers (AT and AE) individually coded the subsequent ten interviews. We reviewed codes as a team and achieved consensus on discrepancies. We also used team meetings to systematically question pre-existing ideas and bias in relation to interview responses. We developed a codebook with specific inclusion and exclusion criteria to help eliminate discrepancies. Examples of responses for both clear and borderline cases were included. After codebook development, JK, AT, and AE individually reviewed all 15 interviews and applied final codes. Representative responses were captured verbatim in Dedoose. We grouped finalized codes into central themes, which we organized into an explanatory framework.

3. Results

We reviewed a total of 6634 charts, which resulted in 9840 visits including 6464 unique patients and 56 providers. Table 1 represents a descriptive summary of all charts reviewed between January 2014 and December 2017. Documentation of contraceptive counseling (39.6%) and contraceptive use (17.9%) remained stable over the time period, but documentation of OKQ rose from 2.6% on initial introduction to 9% after retraining (Table 2).

Table 1.

Descriptive summary of charts reviewed from preventive health visits of reproductive aged females (15-49) between January 1, 2014 and December 31, 2017 at the Community Health Centers, Inc in Salt Lake City, UT, N = 9840.

Variable		Number of charts N(%)
Year	2014	1805 (18.3%)
	2015	2085 (21.2%)
	2016	2827 (28.7%)
	2017	3123 (31.7%)
Contraceptive Counselinga		3888 (39.5%)
One Key Questiona		391 (4%)
Hispanic/Latina ethnicity		8805 (89.5%)
Number of visits per patient	1	6404 (65.1%)
	2	2315 (23.5%)
	3	782 (7.9%)
	4	237 (2.4%)
	5	72 (0.7%)
	6	30 (0.3%)
Provider type	Physician	3375 (34.3%)
	Nurse practitioner	1406 (14.3%)
	Physician assistant	5059 (51.4%)
Provider gender: Female		6453 (0.6%)

^aNumber of charts with documentation present for these variables.

Table 2.

Documentation of contraceptive counseling, One Key Question, and contraceptive use by OKQ time period as represented in the electronic health record Salt Lake City, Utah (US).

Variable	Levels	Pre-OKQ N = 2966	OKQ initial training N = 3728	OKQ retraining N = 3110	p-value
Contraceptive counseling		1233 (41.6%)	1431 (38.4%)	1215 (39.1%)	0.024c
One Key Question		15 (0.5%)	97 (2.6%)	279 (9%)	<0.001a
Contraceptive use		484 (16.3%)	694 (18.6%)	574 (18.5%)	0.030a

OKQ, One Key Question

^aChi-squared test.

3.1. Interrupted time series analysis results

From our model selection procedure, we selected a model with period by date interactions for the OKQ outcome and a date slope with no date by period interactions for the contraceptive counseling outcome. Both models included the additional covariates and random effects described in the methods section. Using smoothed estimates from the interrupted time series analysis model fits, we saw a clear increase in documentation of OKQ particularly after the retraining in March 2017, while a change in rate of contraceptive counseling is less clear (Fig. 1). We have strong evidence for increased odds of OKQ documentation during the initial period of 377 (95% CI: 41–3464), during the initial transition of 2236 (95% CI: 166–30116), and the retraining period of 6197 (95% CI: 243–158309) (Table 3). Additionally, there was strong evidence for a date by transition period interaction for both transition periods representing those strong shifts in slopes of documentation from Figure 1. In addition to the change in OKQ documentation rates at these periods due to practice change, we also have evidence of a decreased odds of 0.68 (95% CI: 0.48–0.96) of documentation, when the patient had a documented contraceptive.

Fig. 1. Interrupted time series analysis:

Smoothed estimated contraceptive counseling and One Key Question rates from the Interrupted time series analysis. Green and Yellow colored zones depict the 6-mo time intervals following the initial implementation and training on OKQ, and the retraining on OKQ, respectively.

Table 3.

Final interrupted time series analysis model fit for One Key Question over the time periods (pre-OKQ: January 1, 2014 - July 31, 2015, initial period: August 1, 2015 – February 28, 2016, and retraining period: January 1, 2017 – December 31, 2017) Salt Lake City, Utah (US).

Random effects estimate:Groups	Name	Std. Dev.	Corr
Patient	Intercept	0.6
Provider	Intercept	1.3
	Date (scaled)	1.06	−0.5
Fixed effects estimates:
Variable	Odds	p value
Intercept	0 (0–0)	<0.001
Documented contraceptive	0.7 (0.5–1.0)	0.03
Nurse practitioner	0.6 (0.2–1.7)	0.4
Physician assistant	0.5 (0.3–1.1)	0.09
Female provider	1.0 (0.6–1.9)	0.9
Visit number	1.0 (0.9–1.1)	0.8
Initial period	377 (41–3464)	<0.001
Retraining period	6197 (243–158309)	<0.001
Initial transition	2236 (166–30116)	<0.001
Retraining transition	4 (0.2–90)	0.3
Initial period: Date (scaled)	1 (0.3–3)	1.0
Retraining period: Date (scaled)	0.2 (0.04–1.3)	0.09
Initial transition: Date (scaled)	13820 (96–1994367)	<0.001
Retraining transition: Date (scaled)	339 (43–2676)	<0.001

Our final model fit for the contraceptive counseling documentation outcome showed no evidence of the OKQ training and addition of the electronic health record field having an effect on the rate of documentation (Table 4). We did find evidence for decreased odds of counseling over time with a multiplicative decrease in odds of 0.72 (95% CI: 0.58–0.89) for each unit increase of the scaled date. Additionally, we found evidence an increased odds of documentation of contraceptive counseling of 1.49 (95% CI: 1.29–1.72) when the patient had a documented contraceptive and a multiplicative increase in odds of 1.36 (1.26–1.46) for each additional visit to the clinic.

Table 4.

Final interrupted time series analysis model fit for contraceptive counseling outcome using the One Key Question time periods (pre-OKQ: January 1, 2014 - July 31, 2015, initial period: August 1, 2015 – February 28, 2016, and retraining period: January 1, 2017 – December 31, 2017).

Random effects estimates:Groups	Name	Std. Dev.	Corr
Patient	(Intercept)	1.2
Provider	(Intercept)	0.6
	Date (scaled)	0.1	−0.7
Fixed effects estimates:
Variable	Odds	p value
Intercept	0.3 (0.2–0.5)	<0.001
Date (scaled)	0.7 (0.6–0.9)	0.003
Documented Contraceptive	1.5 (1.3–1.7)	<0.001
Nurse practitioner	1.2 (0.7–1.8)	0.6
Physician assistant	1.0 (0.7–1.4)	0.97
Female provider	0.9 (0.7–1.2)	0.375
Visit number	1.4 (1.3–1.5)	<0.001
Initial period	1.0 (0.7–1.4)	0.9
Retraining period	1.5 (0.9–2.5)	0.2
Initial transition	1.0 (0.8–1.3)	0.8
Retraining transition	1.2 (0.7–1.8)	0.5

3.2. Qualitative interview results

We interviewed, transcribed, and analyzed transcripts from 15 providers (38%). Eleven are family physicians, and four are physician assistants. The provider perspective on how OKQ affects their contraceptive counseling coalesced into three main domains: patient-level, provider-level, and clinic level and two sub-domains: patient-provider relationship and provider-clinic interface (Fig. 2). While some themes fit a systems-level domain of the effects of OKQ, we did not have sufficient saturation of these themes to confidently draw conclusions. The final model is provider-centric, as opposed to other health systems ecological patient-centric models, as we only interviewed providers. Therefore, each domain linked to the provider’s viewpoint and the presented model fit these data best. The providers also offered multiple limitations of OKQ that we interpreted independently from the main framework.

Fig. 2. Qualitative model:

Model of One Key Question provider-centric domains based on modified grounded theory of community health centers provider interviews. (N = 15).

3.2.1. Patient-level

We defined the patient-level domain as aspects of OKQ that directly affect reproductive autonomy. Themes arising in this domain included access to prenatal care, access to contraceptive methods, and patient pregnancy ambivalence. For one provider, OKQ helped continued access to a method and ensured discussion of options: “[if] they’re on birth control we make sure they’ve got enough supply…or make sure it’s not just ‘oh, I have one more pill, and I’m done.’ So, not leave them hanging. And then if they’re not on birth control we talk about their options” (33-year-old man, physician assistant).

3.2.2. Patient-provider relationship

Themes addressing the aspects of OKQ directly affecting the patient-provider relationship included patient autonomy, OKQ as a bridge to education, language considerations, leading or neutral counseling, personalized care, and consideration of special populations. The way providers used OKQ to perform either leading or neutral counseling offers insight into different approaches to contraceptive counseling. As an example of leading or coercive counseling, one participant remarked, “when it’s [OKQ response] a definite no, I jump straight to long-term reversible [methods]” (37-year-old man, physician assistant). For another provider, the “unsure” response could include leading and neutral counseling:

“If she says that she’s not sure whether or not she wants to become pregnant, we’ll try to let her know that women who are not sure but who aren’t preventing it are just as likely to get pregnant as those who are actively trying. And so, we’ll try to emphasize that…Sometimes people are a little bit more receptive to starting contraception once they realize that that’s something that’s, at least according to the statistics, just as likely to happen to a woman who’s not trying…Sometimes it will lead to a conversation about other issues, like, well, we’re not stopping it [pregnancy]…and her medical history that made it hard for her to conceive, which then can bring up some other underlying [health] issues…which are good to cover anyway.” (39-year-old woman, physician)

However, many providers use OKQ as an open-ended route to discovering more about their patients, supporting their reproductive decisions, and practicing patient-centered care:

“It [OKQ] also has opened up a lot of doors to ironically to find out peoples’ home life. They’ll tell you a lot about the challenges they’re going through with their partner or complications they’re having trying to get pregnant. And so, I’ve had a lot more discussions on infertility…in terms of contraceptive counseling, I think it’s opened up the door to have more conversations…and then we’re able to actually nail down some form of contraceptive for that person and probably wouldn’t have if I hadn’t of asked” (33-year-old woman, physician assistant).

Another provider felt that OKQ has helped them provide “more of my patients with a birth control option that they feel comfortable with” (37-year-old woman, physician). A final example shows how one provider used OKQ to ask further patient-centered questions about method choice: “I sort of ask, ‘Do you have any idea what method of contraception you would like to use?’ if they want some method of contraception. And, I sometimes would say, ‘Have you used anything in the past?’ If they’ve used something in the past, I may direct my counseling towards that, ‘Oh, did you like that? Would you like to use that again? Did that work for you?’” (34-year-old woman, physician).

3.2.3. Provider-level

Themes under the provider-level domain directly affected the provider’s counseling approach and comfort level, acceptability, and frequency of use of OKQ. Themes under this domain included OKQ’s algorithmic approach, ease of OKQ use/implementation, effect of prior clinical experience on use, and indifference to OKQ. Providers remarked that, “It’s [OKQ] just easy…I mean, it’s easier when they are the age … where they can get pregnant…I still think it’s been really helpful” (39-year-old woman, physician assistant). Some providers, however, felt indifferent to OKQ and that it did not change their counseling approach. As one provider noted, “to be honest, I don’t know that it’s [OKQ] changed my own practice too much…I guess my impression was that I had been reasonably diligent about…attention to contraceptive needs…prior to implementing that” (39-year-old man, physician). Another simply answered, “it has not” (37-year-old woman, physician) when asked how OKQ changed their approach to contraceptive counseling.

3.2.4. Provider-clinic interface

We defined this domain as aspects of OKQ that directly affect the provider-clinic interface in providing reproductive health counseling. Main themes included cues to action, improved efficiency, timing of OKQ use during a visit, time constraints, and training on OKQ. Specifically, OKQ served as a reminder to address contraception: “when people have a lot of complex medical questions, it can fall by the wayside. And so, having the One Key Question method and we have like a template that we can drop in, it just helps me be a little more focused for our reproductive-aged patients to make sure I periodically talk to them about contraception.” (34-year-old woman, physician) But others found OKQ would be easier, “if it magically took less time than it does” (39-year-old woman, physician). Time constraints also limited OKQ’s widespread use: “I think that the recommendation for implementation was to ask every woman of childbearing age at every single visit, and I think that practically speaking, that was a little bit too difficult for our practice” (32-year-old man, physician).

3.2.5. Clinic-level

The clinic-level domain included any themes where OKQ directly affected the electronic health record and clinical functioning/workflow around contraceptive care. These themes included electronic health record integration and OKQ implementation in the system. One provider stated, “I have my MAs [medical assistants] ask that question [OKQ]…we have actually built into our templates, so I think I generally kind of ask that question in more or less words, and now it’s asked and documented on all my notes before I’ve even seen the patient” (52-year-old man, physician).

3.2.6. OKQ limitations

Critiques of OKQ, other than time constraints (see section 3.3.4), centered around limitations of the clinic’s electronic health record to provide support tools for OKQ. Specifically, providers wanted integration of educational materials specific to the OKQ response: “We had the actual representative of OKQ come and… they gave us printed handouts, it’s less useful to us than electronic things that we can embed in the EMR…I think if we had electronic things that we could build into our EMR and [make] more routine like other educational material” (38-year-old man, physician). And others pointed out the complete lack of preconception material for those answering “yes” to the OKQ: “if the patient says they don’t want to become pregnant, but they’re not using any method, then it automatically pulls up a list to counsel patients on…But if somebody says they do want to get pregnant, that stuff doesn’t pull up, which is kind of silly” (34-year-old woman, physician).

4. Discussion

This study offers a key perspective on the use of OKQ in preventive health care: the primary care provider. Our mixed-methods study of OKQ implementation by primary care providers found low documented use of OKQ and no change in documented rates of contraceptive counseling. Qualitative data indicated increased attention to patient-centered counseling among the majority of participants, while others employed the OKQ algorithm in a manner that could be considered coercive. But contraceptive counseling is itself complex, and respondents with leading or coercive counseling practices often discussed patient-centered practices as well. Providers appreciated the algorithmic structure of OKQ, its ease of use, and its function as a reminder to address reproductive goals. However, providers voiced indifference about the effect of OKQ on their contraceptive counseling practices and found it difficult to integrate OKQ into preventive health visits due to time constraints and more pressing issues to be addressed. Combining our quantitative findings with provider perspectives, our study suggests limited usefulness of OKQ as a robust screening tool in the primary care setting.

Our findings are at odds with a prior study investigating OKQ’s role in primary care. Stulberg, et al. [13] showed an improvement in patient-reported rates of contraceptive counseling (52% pre-OKQ to 76% post-OKQ, p = 0.04), but similar to a second study by Song, et al. [14] showed no change in patient-reported reproductive (contraceptive or preconception) counseling (69% pre-OKQ to 76% post-OKQ, p = 0.58). Our baseline documentation rate of contraceptive counseling, however, was lower than either of these previous studies (39.7%) despite being in a similar setting of primary care clinics. This was not due to a higher rate of current method use, as our overall documentation of current method use was much lower than national user-reported contraceptive rates (17.8% vs 65.3%) [27]. There are known discrepancies when comparing patient-reported and physician-documented counseling rates [28], and this may account for some of the differences noted above.

A major limitation of this study is relying on providers to document OKQ use, contraceptive counseling, and contraceptive use. Direct, provider-reported use of OKQ and contraceptive counseling instead of relying only on documentation may show improved uptake and counseling respectively. Our qualitative findings, however, do demonstrate low overall uptake as many remarked that OKQ did not change their approach to contraceptive counseling. Additionally, we were unable to ascertain from chart review the consent process for performing counseling and thus, low rates of documentation could also reflect actual practice if patients declined counseling when offered. Further, OKQ training is now managed by Power to Decide and requires a longer, standardized online training than what participants received in 2015. It is possible that use of OKQ would be higher with the current training model. Limitations of the interrupted time series analysis framework include assumptions of a pre-existing linear trend in contraceptive counseling that would continue without the intervention (OKQ retraining), and the possibility of time-varying confounders not accounted for in the model. There are many strengths to the study including the use of interrupted time series analysis, a large number of data points for the interrupted time series analysis, and investigation of primary care providers’ perspectives on OKQ.

A simple screening tool for reproductive goals counseling is needed within primary care settings, but as evidenced in this study OKQ may not be the answer. Whereas OKQ may be useful in some clinical settings, it may not be in others. Our conclusions conflict with the findings of prior OKQ studies utilizing patient-reported data, and in this setting, primary care providers used the OKQ infrequently and identified limitations. The tool’s major advantage is brevity. As the name implies, it is quick to use. With its lack of nuance, alternatives to OKQ have been suggested [16]. However, we still remain in search of an instrument that will consistently be used by primary care providers in a variety of settings, be appreciated by those receiving care, and will improve outcomes.

Abbreviations:

OKQ

One Key Question