Cervical Cancer Screening in a Sexually Transmitted Disease Clinic: Screening Adoption Experiences From a Midwestern Clinic

Abstract

Objectives. We examined whether a sexually transmitted disease (STD) clinic could reach women who had not received a Papanicolau (Pap) test in the past 3 years. We also explored staff attitudes and implementation of cervical cancer screening.

Methods. Women (n = 123) aged 30 to 50 years were offered cervical cancer screening in an Indiana STD clinic. We measured effectiveness by the patients' self-reported last Pap test. We explored adoption of screening through focus groups with 34 staff members by documenting their attitudes about cervical cancer screening and screening strategy adaptation. We also documented recruitment and screening implementation.

Results. Almost half (47.9%) of participants reported a last Pap test 3 or more years previously; 30% had reported a last Pap more than 5 years ago, and 11.4% had a high-risk test outcome that required referral to colposcopy. Staff supported screening because of mission alignment and perceived patient benefit. Screening adaptations included eligibility, results provision, and follow-up.

Conclusions. Cervical cancer screening was possible and potentially beneficial in STD clinics. Future effectiveness-implementation studies should expand to include all female patients, and should examine the degree to which adaptation of selected adoption frameworks is feasible.

Cervical cancer morbidity and mortality rates have remained stable in the United States for the past several years,^1,2 with disparities rooted in sociodemographic and health system characteristics. These disparities are well documented and indicated by the lack of screening and follow-up for abnormal Papanicolau (Pap) tests.^3–6

The few studies with systems implications for cervical cancer screening observed reductions in screening, colposcopy, or survival disparities when the health system and services barriers were removed. Doyle et al. found that providing breast and cervical cancer screening in an urban walk-in medical clinic increased access to follow-up colposcopies.⁷ Farley et al. found that equal and unfettered access to cervical cancer screening in a military care system eliminated differences in survival rates for women with cervical carcinoma.⁸ Castle et al. observed that cervical cancer screening increased among underserved, rural Mississippi women when they were offered the opportunity to self-collect samples for human papillomavirus (HPV) testing compared with completing free Pap testing at a clinic site.⁹ Finally, Hitt et al. found that tele-colposcopy services increased access to follow-up for abnormal Pap tests among women in rural Arkansas.¹⁰ These studies are encouraging and lead the way for additional system-related studies to reduce cervical cancer screening and follow-up disparities. Venues that can provide high-risk HPV (hrHPV) screening access and navigation to colposcopy will be critical to cervical cancer prevention efforts as diagnostic strategies evolve.

The sexually transmitted disease (STD) clinic is an example of a system opportunity that may increase access to cervical cancer screening. STD clinics provide services at low or no cost to the patient, and are safety net providers of care for racially/ethnically diverse populations who are likely under- or uninsured and have tenuous relationships with the traditional health care system.^11–13 These characteristics have been associated with reduced utilization of cancer screening.^14–17 Further, recent Centers for Disease Control and Prevention STD policy guidelines¹⁸ have highlighted the importance of STD clinics for cervical cancer reduction because the prevalence of hrHPV among female STD clinic patients ranges from 27% to 42%.^19,20

Two recent studies found that some STD clinics provide Pap or HPV testing, although the extent of cervical cancer screening in STD clinics generally remains unknown.^21,22 Although STD clinics might be likely venues for cervical cancer screening, the central challenge remains one of organizational–cultural orientation as STD clinics are epidemiologically oriented and provide episodic care, screening, and follow-up primarily for HIV, syphilis, gonorrhea, chlamydia, and trichomoniasis. The persistent strains of a sexually transmitted infection (HPV) cause cervical cancer, yet it is not clear whether STD clinics could provide cervical cancer screening, and if they could, whether these services would contribute to the reduction of health disparities. With health care costs in mind, the ability of a clinical setting to provide cervical cancer screening is not sufficient. What must be demonstrated is that such a setting actually reaches women who are underscreened, or by definition have “not had a Pap test in at least 3 years.”²³

We examined the effectiveness of STD clinic-based cervical cancer screening in reaching women who had not received a Pap test in the past 3 years, and explored the clinical implementation of cervical cancer screening in this setting. Our study had 2 goals: innovation strategy effectiveness and implementation. We used an effectiveness–implementation hybrid type 1 design, which allowed for the testing of this innovation strategy while gathering information on its adoption and delivery.²⁴ Thus, our primary research aim was screening effectiveness, and our secondary aim was the adoption and implementation of cervical cancer screening. This design reflected the need to shorten the long delay between innovation and translation in public health settings.^25,26

The multilevel adoption framework of Frambach and Schillewaert^27–29 provided the primary conceptual orientation for the implementation study aim. This framework had sufficient construct flexibility that allowed for varied approaches to measuring adoption elements, and for adaptation in real-world contexts.³⁰ In addition, like other multilevel adoption frameworks,³¹ it focused on both organizational and individual characteristics. The framework informed variables selection for the implementation study component, but because of the exploratory nature of this study, all framework components were not operationalized. Selected constructs included staff receptiveness and readiness for the innovation, attitudes, and opinions about cervical cancer screening as an innovation, innovation alignment or fit with the clinic mission, trialablity, and problem solving around encountered barriers. Two additional constructs were integrated into the study from other models because of their close alignment with our primary framework. They included innovation characteristics of complexity and adaptability,³² and the perceived advantage of innovation.³³

Our specific study aims were (1) to determine whether offering cervical cancer screening in an STD clinic venue would reach women who needed it, as measured by reported last Pap smear, and (2) to explore the STD clinic’s adoption and implementation of cervical cancer screening as measured by staff attitudes and observed screening implementation.

METHODS

Our study was conducted at the Bell Flower Clinic in Indianapolis, Indiana, a walk-in county health department STD clinic. This clinic served 4063 women, with 5266 clinic visits in 2013. The patient population was 38% women, of whom 33% were aged 30 to 44 years; 13% were aged 45 years or older. Approximately 57% of the patients were Black, 34% were White, 7% were Hispanic, and 2% were of other, undefined racial/ethnic backgrounds. The majority of patients were examined by a clinician, and fewer than one third (29%) of clinic encounters were “express” or “fasttrack” visits, in which asymptomatic patients with no known partner with an STD gathered their own sample for STD testing without examination by a clinician.^34,35

This clinic has 53 staff members, with 30 clinicians (7 physicians, 2 physician assistants, 6 nurse practitioners, 5 registered nurses, 7 staff nurses, and 3 clinical aides), 15 field staff (10 disease intervention specialists and 5 surveillance and data staff), and 8 administrative staff who provide clerical and registration support. The clinic director is 1 of 3 infectious disease physicians. Most of the staff (83%) are full-time employees, although the 9 part-time employees are all clinicians (5 physicians and 4 nurse practitioners).

Participants and Procedures by Aim

To evaluate effectiveness, cervical cancer screening was offered to female clinic patients aged 30 to 50 years who were triaged for clinician examination between June 2012 and January 2014. Patients were recruited via waiting room posters in the clinic that advertised free cervical cancer screening and by clinic staff invitation to eligible women at triage. Cervical cancer screening included a Pap test with HPV DNA testing. A brief enrollment survey collected participant self-reported demographic characteristics, cervical cancer screening history, whether they had a regular doctor, and their preferred contact method for screening results. Participants were given a $20 gift card for completion of the enrollment survey, and a pamphlet about HPV and cervical cancer.

The HPV DNA testing used the cobas4800 HPV assay (Roche Molecular Diagnostics, Indianapolis, IN). Following clinic routine, duplicate (2 swabs in a single collection device) self-obtained vaginal swabs were collected using the BD Culturette II (BD Diagnostics, Sparks, MD) and transported to the laboratory for routine chlamydia and gonorrhea screening. Cervical samples were collected using a broom device eluted into a liquid-based cytology medium (Hologic, Marlborough, MA) in accordance with the cobas package insert. HrHPV testing on the cobas provided positive or negative results for HPV16, HPV18, and a pool of 12 other high-risk HPV types.

Three attempts were made to contact all participants with results using participant-preferred methods. Those who were hrHPV+ and who had a Pap result of atypical squamous cells of undetermined significance were navigated to colposcopy. Study materials and all communications were offered in English and Spanish.

To explore screening adoption and implementation, we conducted 60-minute focus groups before and after screening implementation to gather staff attitudes and opinions about the innovation. We limited focus group eligibility to full- and part-time administrative, clinical, and surveillance or field staff members. We limited screening implementation eligibility to registered nurses, nurse practitioners, physician assistants, and physicians. We sent invitation emails to all full-time and part-time STD clinic staff to participate in the focus groups. We separated staff members who volunteered for the focus groups into groups by clinic role (clinical, field or surveillance, administrative) to facilitate free expression about the adoption of cervical cancer screening in the clinic. As guided by the Frambach and Schillewaert framework, focus group questions solicited attitudes and opinions about cervical cancer screening in STD clinics generally, in this particular clinic, and advice about the implementation of the proposed screening strategy. The final focus group, which occurred after the completion of screening implementation, presented screening effectiveness data, and repeated the initial focus group questions soliciting attitudes about the innovation implementation.

Organizational observation of screening implementation focused on recruitment patterns measured by dates of screening, staff provision of screening tracked by name and clinic role, and barriers to implementation, measured by observed or reported protocol deviations. Field notes focused on screening provision (by date and staff member), and protocol deviation.

We used a mixed methods design, with sequential data gathering followed by simultaneous data gathering.³⁶ Methods were complementary, because the qualitative data informed the contours and the implementation of the innovation strategy.³⁷ Thus, qualitative data gathering occurred with the initial staff focus groups. Quantitative and qualitative data were simultaneously gathered during the screening implementation. Finally, qualitative data were gathered at the conclusion of the screening implementation with the final staff focus groups.

Statistical Analyses

Variables of interest to evaluate screening strategy effectiveness included sociodemographic characteristics (age, race/ethnicity), health system access (being insured, having a regular doctor, time of last Pap test), cervical cancer screening outcomes, and whether the patient was reached for results provision. Data sources included an enrollment survey, laboratory report, and tracking of the patient results contact. We measured effectiveness by whether participants were underscreened or had not undergone a Pap test in at least 3 years. A secondary measure of effectiveness was the extent to which screening identified women in need of follow-up colposcopy.

Screening data involved both scaled and dichotomous measures, and included testing outcomes, demographic characteristics, study uptake, and availability to receive results. Univariate descriptions were followed by bivariate investigations of associations among the variables with a χ² test of difference (P < .05) among participant sociodemographic characteristics, health system access characteristics, and testing outcomes.

Focus group data were transcribed and then coded independently by 2 separate investigators using an a priori approach guided by key framework elements of (1) innovation acceptance, (2) mission fit, (3) perceptions of peer and patient acceptance, and (4) issues with adaptability. We held a verification conference to establish the primary coding scheme, followed by resorting using axial coding based on the adoption constructs.³⁸ After the observation that there were no theme differences among staff groups from the initial focus group sessions, we decided to collapse the focus group data for aggregate evaluation. Recruitment data included dates of patient screening and the staff member listed on the consent form. Dates were visually displayed on a timeline for pattern observation. Implementation evaluation included counting the number of times an individual staff member screened patients or engaged in follow-up navigation to colposcopy, and calculating the share of screening completed by that specific staff member. The same analysis was repeated based on staff role participation by specific job (physician assistant, nurse practitioner, and so on) and role (clinician, administrative, and so on).

RESULTS

Our study involved 123 patients who received cervical cancer screening and 34 STD clinic staff who participated in focus groups and implemented screening during the study period. Screening participants reflected the racial/ethnicity distribution of female Bell Flower Clinic patients: 56.1% were non-Hispanic African American, 35.8% were non-Hispanic White, and 6.5% were Hispanic. The mean age of participants was 37.9 years (SD = 5.1; range 30–50 years). The bivariate associations among age, race/ethnicity, and other variables were not significant.

More than half (64%) of the clinic staff participated in the focus group discussions before and after the screening implementation (Table 1).

TABLE 1—

Focus Group Participation by Type of Staff: Indianapolis, IN, 2013

Variable	Total Staff	Focus Group Participation	% Participating
Clinicians
Infectious disease physician	2	1	50
PT physician	5	0	0
Physician assistant	2	2	100
Nurse practitioner	2	2	100
PT nurse practitioner	4	3	75
Registered nurse	5	5	100
Staff nurse	7	0	0
Field/surveillance
Disease intervention specialist	10	8	80
Surveillance and data management	5	3	60
Administrative
Administrative	8	8	100
Clinical aidea	3	2	67
Total	53	34	64

Note. PT = part-time.

^aIncludes routers, triage, and translators.

Preimplementation and Adaptation of Screening Strategy

Staff opinions about cervical cancer screening in the STD clinic before implementation were positive and focused on alignment with the STD clinic mission and the likely patient benefit. Staff perceived that patients were largely uninsured and, therefore, did not have regular medical care, and that the STD clinic was likely their only primary care location. Some administrative staff stated, “Women often call and ask if we do Paps, but they are really looking for good medical care, ’cause that’s what it (Pap smear test) means to them.”

An anticipated primary implementation challenge was reaching patients with screening results and to navigate them to colposcopy if needed. Staff experience was that patients were generally not available for results. One clinician said, “You won’t reach them (with results). One time, gone. Been here one time and you don’t see them again.”

Focus group themes by construct are shown in Table 2. Differences in opinions about the clinic’s adoption of cervical cancer screening among the staff groups (administrative, clinical, and field or surveillance) were not observed.

TABLE 2—

Organizational Adoption Factors and Outcomes (n = 34): Indianapolis, IN, 2013

Adoption Factor	Themes
Preimplementation
Opinions about cervical cancer screening in STD clinics generally and in this clinic	Needed by patients because they are uninsured and do not get medical care elsewhere.
	HPV is an STD.
	Feasible. Some clinicians have Pap experience.
	Might benefit STD clinic later with new funding.
Opinions about screening methods	Should do both Pap and HPV. Patients value Pap smears, and co-testing is standard of care.
Opinions about challenges offering screening in this STD clinic	Belongs here but concerned about follow up—reaching and finding patients, getting patients to colposcopy.
	Should screen women younger than 30 y, especially in STD clinics.
	Lack of Pap literacy among patients—they think they have a Pap when they get a pelvic examination.
	Balancing a women’s study with other studies and clinical work.
	Results: reaching women, giving negative/concerning results.
Perceptions of clinic peer opinion about benefits and challenges of cervical cancer screening in this clinic	Some may not adjust well to change in the clinic practice given current clinical burden and balance of research and clinical services.
Perceptions of professional STD peer opinions (outside clinic) about benefits and challenges of cervical cancer screening in this clinic	Way to keep STD clinics open and financially viable.
	Rural clinics will have a more difficult time with laboratory and follow-up infrastructure.
Perceptions of patient response to offer of cervical cancer screening and results contact	Patients want it. They request it. But really want the Pap test.
Adaptability of innovation	Several recommendations for intervention adaptation to better “fit” the clinic (Figure 1).
Postimplementation
Opinions about the challenges with cervical cancer screening implementation	Age criteria too strict for clinic population.
	FastTrack patients could not participate, but should.
	Triaging Pap test patients—very difficult given staff resources.
	Competition among research projects—hard to balance them and also clinical services.
	Consenting process (research) takes too much time.
	Patient lack of Pap literacy.
Opinions about the necessary improvements if implementing cervical cancer screening in clinic in future	Expand age criteria to include women aged ≥ 18 y.
	Establish a more efficient Pap triage method in the clinic.
	Reflex pap test HPV+ fast track patients.
	Integrate medical records to document last Pap.
	More clinical time for consultation with patients.
	Greater focus on navigation and include insurance.
	Consider integrating colposcopy in the clinic.

Note. HPV = human papillomavirus; Pap = Papanicolau; STD = sexually transmitted disease.

The preimplementation focus group discussions also involved the proposed implementation model for cervical cancer screening in this particular STD clinic. As shown in Figure 1, 5 adaptations were made to the innovation strategy during the focus group sessions.

FIGURE 1—

Intervention adaptations from preimplementation focus groups: Indianapolis, IN, 2013.

Note. DIS = disease intervention specialist; hrHPV = high-risk human papillomavirus; Pap = Papanicolau.

During the first set of focus groups, staff noted that adaptations improved their perceptions of screening trialablity because they addressed several clinic realities that could have served as barriers to implementation. For example, the shift of results provision from the field staff to the research staff removed staff concerns about reporting cervical cancer screening results to patients. Even at this clinic, with experienced clinical and field staff who often provided results consultation following HCV and HIV testing, the concern was about patient reaction and staff management of reactions. Clinic staff perceived the results consultation component of the innovation to be too difficult for them at this stage of adoption, and some field staff explained, “I am concerned about the logistics around giving people results. I think they will freak. Anything you talk about in terms of cancer, they usually freak out. So, not sure about that.”

Screening Strategy Effectiveness

Screening participants appeared to need cervical cancer screening because the majority (78%) reported not having a regular doctor. Of the 56 women who reported their insurance coverage, 75% reported not having insurance (Table 3). Of the 94 women who reported their last Pap test, 47.9% reported not receiving a Pap test in at least 3 years. The 30% of women who reported no Pap in at least 5 years were more likely not to have a regular doctor. The associations between reported last Pap test and age or race/ethnicity were not significant. Unfortunately, not all participants (n = 123) reported all data. Missing values analysis was difficult based on the small sample size and the lack of data elements that could assist particular elements (insurance, last Pap test). The most robust of health systems data were reported regular physician followed by the reported last Pap test (n = 94), and then followed by reported insurance (n = 56).

TABLE 3—

Reported Papanicolau Test History by Regular Doctor (n = 94): Indianapolis, IN, 2013

Reported Last Papanicolau Test	No. (%)	No Regular Doctor OR (95% CI)
This year	3 (3.2)	0.86** (0.73, 1.00)
1–2 y ago	46 (48.9)	0.59 (0.22, 1.54)
3–5 y ago	16 (17)	1.30 (0.36, 5.40)
> 5 y ago	28 (29.8)	3.30* (0.91, 12.40)
Never	1 (1.1)	. . .
Total	94

Note. CI = confidence interval; OR = odds ratio.

*P ≤ .05; **P ≤ .01.

HrHPV positivity among the sample of patients mirrored previous findings in this setting.²¹ Of the participants, 23.6% were positive for hrHPV, 5.7% were positive for hrHPV 16 or 18, and 18.7% were positive for other high-risk types of HPV (some or all of the following: 31, 33, 39, 45, 51, 52, 56, 58, 59, 66, and 68), and 11.4% of participants tested positive for hrHPV and had a Pap outcome of atypical squamous cells of undetermined significance. Fewer than 10 women were hrHPV-positive with Pap outcomes of low-grade squamous intraepithelial lesions (6.5%). No statistically significant differences by race/ethnicity or age and testing outcomes were found. The associations between reported last Pap test and cervical cancer screening outcomes were not significant.

When offered an opportunity to obtain a cervical cancer screening test in the context of the STD clinic visit, 99% of eligible clinic patients offered screening by their clinician consented to be tested. Despite staff concern about patient results availability, 91.9% of participants were available when contacted about their cervical cancer screening results. Patient acceptance of screening was notable in view of the fact that recruitment required 7 months longer than the initially estimated reported clinic census.

Postimplementation

Postimplementation focus group data revealed several challenges that likely contributed to the pace of recruitment. Pap test triage was difficult because of the lack of clinicians with this expertise. Further, the screening was itself a research study that required consent procedures, and the staff felt that this was a primary barrier to screening implementation. One clinician stated,

I came from a clinic where I easily did 20 Pap smears a day. . . . It is not going to take this long because you’re used to doing it. But in general we all feel like it slows us down to do a research study because you don’t do it often enough and then you’ve got more than one and you’re trying to figure out what paper work… and I’ve got to collect this swab in the correct order and record it here so that I can get this gift card so I can give it to them. It slows you down. If it is a routine thing it would flow and we’d catch more women all around.

Although screening refusals were tracked by clinicians who offered cervical cancer screening to their patients, the clinic did not track the number of women who might have asked to be screened based on poster recruitment. A closer look at recruitment revealed that 5 of the clinical staff (2 nurse practitioners, 1 registered nurse, and 2 physician assistants) provided 83% of the screening. Screening implementation among them was fairly consistent throughout the study period, with the exception of 1 physician assistant who began to offer screening during the last 6 months of the study and 1 nurse practitioner who was hired to replace an outgoing colleague (Table 4).

TABLE 4—

Distribution of Implementers Across Staff Function: Indianapolis, IN, 2013

Variable	Total Staff	Implementers	% Potential Implementers	% Participants Served
Clinicians
Infectious disease physician	2	2	100	13a
PT physician	5	0	0
Physician assistant	2	2	100	29
Nurse practitioner	2	2	100	32
PT nurse practitioner	4	1	25
Registered nurse	5	3	60	37
Staff nurse	7	0	0
Field/surveillance
Disease intervention specialist	10	0	NAb
Surveillance and data	5	0	NAb
Administrative
Administrative	8	1	13	11.8a
Clinical aide	3	0	0
Total	53	11		NA

Note. NA = not applicable; PT = part-time.

^aOne staff member in this role provided clinical consultation and navigation to colposcopy for women with abnormal test outcomes.

^bRole was removed from adapted intervention.

Despite recruitment challenges, staff opinion about the fit of cervical cancer screening in this clinic environment continued to be positive (Table 2). Staff also recommended that an STD clinic should not limit cervical cancer screening to women aged 30 years or younger because women of all ages have sexually transmitted infection exposures and are underserved by the medical system. One clinician said,

I’m thinking the first girl I remember that we had that actually ended up having carcinoma in situ was 17 years old! That concerns me now because they’re not even advising Paps for girls under the age of 18 I don’t think. . . . And so they’re gonna have plenty of opportunity to have those years go by where they’re developing problems and not clearing it.

DISCUSSION

Our effectiveness-implementation study yielded several observations related to cervical cancer screening in STD clinics. First, if an STD clinic offered cervical cancer screening to their patients, it was likely that patients would need it. The reasons for this were not entirely clear, as study participants might have volunteered because they knew they needed a Pap test. This study limitation notwithstanding, it was likely that because of similarities in health system access to STD clinic patients elsewhere, these STD patients were underscreened for cervical cancer because the majority of participants did not have a regular doctor and many did not have insurance.

Second, the exploration of adoption and implementation of screening in this clinic suggested that an STD clinic could implement cervical cancer screening, although attention should be paid to strategy adaptation. As noted by Frambach and Schillewaert, trialability had a positive relationship with adoption. We observed that innovation adaptation before implementation improved staff opinion of trialability, because it reduced the risk associated with early adoption. Staff concerns about providing cervical cancer screening results to patients were addressed through adaptation, and therefore, risk of adoption was reduced. Further, staff focus group data before and after implementation clearly indicated supportive beliefs toward the innovation. Presenting screening effectiveness data to staff during the final focus group discussions appeared to reinforce the already positive attitude toward the innovation, because the data reinforced staff beliefs that the innovation was needed by their patients.

We speculate that environmental influences, such as financial resources in the form of billable time or cost reimbursement, might facilitate greater staff adoption of screening. However, that we had 4 early adopters who provided 83% of the screening pointed to individual characteristics. This was consistent with the suggestion by Frambach and Schillewaert that perceived innovation characteristics might mediate environmental influences.

Because 47.9% of participants reported a last Pap test 3 or more years ago, the public health implication of implementing cervical cancer screening in an STD clinic is a potential decrease in screening disparities, especially among women who are uninsured and otherwise underserved by the health care system.

We selected the Frambach and Schillewaert framework because of its construct flexibility and multilevel (individual and organizational) focus³⁰; however, we did not full apply the framework in this exploratory study. We integrated some aspects of the framework. The question remains whether this framework is the most appropriate to guide future studies, and if so, how researchers could demonstrate that such framework adaptation is appropriate and robust.

The option for staff implementers to collaboratively adapt the screening strategy before implementation might also be an important contributor to implementation success. This co-creation process involving the implementers as innovation developers was similar to the collaboration impact noted elsewhere,³⁹ and might be part of the “secret sauce” of success for implementation.

As has been demonstrated elsewhere, cervical cancer screening is cost effective,⁴⁰ especially when offered to underscreened women. The effort to identify venues that can reach underscreened women could make an important contribution to reducing rates of cervical cancer in the United States within the current resource-constrained environment.