Abstract
Objective:
Assess the impact of a multicomponent intervention in women with cervical dysplasia who were treated with loop electrosurgical excision procedure (LEEP), as well as the time between colposcopy and treatment.
Design:
Retrospective cohort study.
Intervention:
Clinic participation in a multicomponent cervical cancer prevention program that included community outreach, patient inreach and navigation, as well as provider capacity building with in-person training and ongoing telementoring through Project ECHO.
Main Outcome Measures:
Medical records were reviewed to assess women with cervical dysplasia undergoing treatment with LEEP within 90 days of colposcopy, as well as time between colposcopy and treatment. Baseline data from Year 1 were compared with each subsequent year of implementation. Additional variables examined included patient’s age, previous history of abnormal screening results, percentage of families living below poverty line based on county of residence, parity, and clinic site. We performed logistic regression and multiple linear regression to assess the programmatic impact in the outcomes of interest by year of program implementation.
Results:
A total of 290 women were included in the study. The proportion of women undergoing treatment within 90 days of colposcopy increased from 76.2% at baseline to 91.3% in Year 3 and 92.9% in Year 4 of program implementation. The odds of undergoing treatment within 90 days were 5.11 times higher in Year 4 of program implementation compared with baseline. The mean time between colposcopy and LEEP decreased from 62 days at baseline to 45 days by Year 4 of program implementation.
Conclusions:
Implementation of our multicomponent cervical cancer prevention program increased the proportion of women undergoing LEEP within 90 days of colposcopy and decreased the time between colposcopy and LEEP. This program has the potential to support cervical cancer prevention efforts and could be implemented in other low-resource settings.
Keywords: Cervical cancer prevention, cervical dysplasia treatment, LEEP, multicomponent intervention
INTRODUCTION
Background
Cervical cancer is a preventable disease caused by persistent infection with high-risk human papillomavirus (HPV). [1, 2] Considering that cervical cancer develops over an extended period of time [3] and that in high-income countries like the U.S., there are existing strategies for primary and secondary prevention of cervical cancer that are highly effective, no woman should die from cervical cancer. The HPV vaccine has been available in the U.S. since 2006 and has been shown to be safe and highly effective in preventing HPV infection, high-grade cervical dysplasia, and cervical cancer. [4–8] Additionally, screening is widely available for cervical preinvasive disease using cytology and/or HPV testing. [9, 10] Lastly, there are effective treatments [11–13] for women identified with pre-cancerous lesions such as excision with loop electrosurgical excision procedure (LEEP). The secondary prevention strategies that treat preinvasive disease provide protection against progression to cervical cancer when performed at the appropriate ages and intervals according to national guidelines. [14, 15]
Despite available preventative strategies and tools, cervical cancer rates still remain high in low-resource settings including medically underserved areas (MUA) of the U.S. [16, 17] There are ethnic and racial disparities in the incidence and mortality rates of cervical cancer. For example, the highest rates of new cervical cancers nationally are among American Indian and Alaska Native women (10.8 per 100,000 people, 2014–2018) followed by Hispanic women (9.6 per 100,000 people, 2014–2018). [18] The highest mortality rates of cervical cancer nationally are among non-Hispanic black women (3.4 per 100,000 people, 2015–2019), followed by the American Indian and Alaska Native group (3.1 per 100,000 people, 2015–2019) and Hispanic women (2.5 per 100,000 people 2015–2019).
Women with abnormal cervical cancer screening/diagnostic results, who live in MUAs, may have access barriers to treatment, especially if they live in areas considered health professional shortage areas [19], or if they lack health insurance. An average of 8.6% of the US population is uninsured, though the rates are much higher in certain states, such as Texas, where 18.4% of the population is uninsured. [20, 21] Additional barriers to treatment of cervical preinvasive disease include lack of patient awareness and lack of patient knowledge [19, 22], lack of adequate system infrastructure to identify and navigate patients in need of treatment, lack of clinical infrastructure to perform diagnostic and treatment procedures as well as regional lack of trained professionals delivering these services. [23–25]
In 2018, Dr. Tedros Adhanom Ghebreyesus, Director General of the World Health Organization (WHO), released a call for action to eliminate cervical cancer worldwide, and in 2020 this call for action was adopted as the global strategy for cervical cancer elimination by the World Health Assembly. [26] The aspirational goal of this initiative is to eliminate cervical cancer as a public health problem by 2030 through implementing initiatives to achieve three targets: 1) vaccinate 90% of girls with the HPV vaccine; 2) screen 70% of women for cervical cancer using a high-performance test; and 3) treat 90% of women with cervical pre-cancer and cancer.
The target for treating 90% of women with pre-cancerous lesions and cancer may be challenging to accomplish for several reasons including: socioeconomic status of women (insurance, language, legal status, lack of transportation barriers), lack of access to care (fewer places to receive services) and lack of providers trained to deliver these services. [24] Multicomponent interventions to address the aforementioned barriers are needed in order to reach the treatment targets from the WHO elimination initiative.
The MD Anderson Program for Reducing Cervical Cancer in Texas (PRCC-Texas) is a multicomponent intervention to address barriers to cervical cancer screening, diagnosis, and treatment. [23] The PRCC-Texas Program was initially implemented in two Federally Qualified Health Center (FQHC) systems and one mobile health clinic in the Rio Grande Valley (RGV) along the Texas-Mexico border. Su Clinica is a FQHC serving uninsured and underinsured patients in three locations: Brownsville and Harlingen (both in Cameron County) and Raymondville (Willacy County). The Dysplasia and Cancer Stop Clinic provides screening, colposcopy and LEEP services for women in McAllen (Hidalgo County). The UTHealth McGovern Medical School Mobile Health Clinic, staffed by a Physician Assistant and a lay health worker (LHW) provides basic prevention services for uninsured women in Cameron County. The Su Clinica system was the first to implement and adopt the program. They hired navigators to provide inreach and community outreach and their providers participated regularly in Project ECHO sessions. The McAllen and mobile van clinics joined the program later and did not hire navigators to provide community outreach and education. Their providers also attended Project ECHO sessions.
The multicomponent PRCC-Texas program includes community outreach and education (for screening demand generation), navigation services to identify and educate women in need of screening/treatment, navigation to care when evaluation and treatment is required, access to diagnosis (colposcopy) and treatment (LEEP) services, training, and education of providers to offer colposcopy and LEEP services and programs to support on-going provider training. The education/navigation of patients was performed by trained bi-lingual navigators, and the provider training was performed by specialists from the MD Anderson program in-person and virtually using the Project ECHO® telementoring model. Our navigation approach focused on hiring/training navigators to perform outreach (in the community) and in-reach (within the clinics) education and navigation. Our navigators focused on performing in-reach to women with previous abnormal screening results or women past-due for screening or with unknown screening history.
The details on the program have been previously described. [23] This multicomponent program has been shown to have a significant impact in increasing cervical cancer screening rates and increasing the proportion of patients with abnormal screening results who receive follow up evaluation with colposcopy and treatment with LEEP, as appropriate, at three participating clinic systems in the RGV. [23]
Objectives
The overall goal of this study was to assess whether the PRCC-Texas program (by year of program implementation) impacted the proportion of women with high-grade cervical precancerous lesions (cervical intraepithelial neoplasia grade 2 and/or 3 [CIN 2/3]), who underwent treatment with LEEP. Furthermore, we evaluated the time from colposcopy of CIN 2/3 to treatment with LEEP.
METHODS
A retrospective cohort study was performed of prospectively collected data from women who underwent cervical cancer screening at as part of the PRCC-Texas from November 2014 to November 2018. Institutional Review Board (IRB) approval was obtained from The University of Texas School of Public Health (protocol #HSC-SPH-20–0577) to perform this analysis. A waiver of informed consent was granted for this study by the same IRB. All the steps/methods were performed in accordance with the relevant guidelines and regulations. As part of the PRCC-Texas program, all data were collected by patient navigators and trained data collection personnel and retrospectively reviewed and analyzed for this study. We assessed if the exposure of interest (year of implementation of the PRCC-Texas Program (Year 1 (baseline), Year 2, Year 3, Year 4) had an impact on the proportion of women diagnosed with CIN 2/3 who underwent appropriate and timely treatment with LEEP. Baseline data included any patients with data collected before or during Year 1 of the program (November 2014-October 2015). Each subsequent implementation year was compared with this baseline year through November 2018.
The outcomes of interest were defined as: 1) the proportion of women with CIN 2/3 undergoing LEEP within 90-days of colposcopy; and 2) the number of days between colposcopy/biopsy and LEEP. These outcomes were selected based on the Centers for Disease Control and Prevention (CDC) National Breast and Cervical Cancer Early Detection (NBCCED) program recommendations which include that >90% of women with high-grade cervical dysplasia undergo treatment, and that >80% of these women receive treatment within 90 days of colposcopy. [27–29]
The medical records were reviewed for cervical screening, diagnosis, and treatment results. Patient level covariates included age, smoking status, previous history of abnormal screening and parity as these covariates have been identified in the literature as being predictors of timely diagnosis and treatment of patients with abnormal screening results. [19, 30] We also included patients’ zip code of residence as a proxy to socioeconomic status (SES) using the proportion of families living below poverty level. To code the poverty level variable, we reviewed the zip code home address of all included patients and categorized them into quartiles previously set for the RGV for families living below poverty level between 2014–2018 by zip code.
Data handling and record keeping
Study data were managed using the Research Electronic Data Capture (REDCap) electronic data capture tools hosted at MD Anderson. REDCap (www.project-redcap.org) is a secure, web-based application with controlled access designed to support data capture for research studies, providing: 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless downloads to common statistical packages; and 4) procedures for importing data from external sources. [31]
Statistical methods
Descriptive statistics were performed for each of the variables of interest as a total and per year of program implementation. Additionally, bi-variate analysis was performed to identify variables that could be associated with the outcome of interest and year of program implementation. A stepwise approach was then used to select the variables that best fit the model in a multiple logistic regression for the first outcome (LEEP appointment within 90 days of colposcopy) and multiple linear regression for the second outcome (days lapsed between colposcopy and LEEP). Significance levels were set at p-values < than 0.05. We used Hosmer-Lemeshow test to assess the goodness of fit of the final model. The data were analyzed using STATA 17.0. [32]
RESULTS
As shown in Figure 1, 14,846 women underwent screening as part of the PRCC-Texas during the study period (2014–2018). A total of 12,760 women had normal/benign results (including cytology showing Atypical Squamous Cells of Undetermined Significance (ASC-US) with negative HPV testing) and did not require further evaluation. An additional 94 women had missing cytology date/results and were also excluded. A total of 1,992 women had abnormal results and underwent colposcopy/biopsy. Of these, 1,514 had normal/benign or low-grade (CIN 1) biopsy results not requiring further treatment and were excluded from further analysis. An additional 140 women were excluded because they did not have a biopsy at the time of colposcopy (n=110) or had missing biopsy results (n=30). Of the remaining 338 women, an additional 49 women were excluded due to a diagnosis of adenocarcinoma in-situ (AIS) (n=4) or invasive cancer (n=23) or were referred for hysterectomy (n=20) or cold knife cone (CKC) (n=1). The remaining 290 women were referred for LEEP and comprise our study population.
Figure 1.
Study flow chart
The demographic data for the 290 women included in this study are presented in Table 1. The mean age at diagnosis was 34 years (median of 35 years and range of 20–65 years). The majority of patients were non-smokers (86.7%), with 80.6% never smokers and 6.5% previous smokers. Almost half (47.3%) of the women had a previous history of abnormal screening results prior to participating in the PRCC-Texas program. Most of the women (67.0%) lived in zip codes with 25–58% of families living below the poverty line. The majority of participants (83.3%) had a history of two or more births and of these, 31.1% had more than four births.
Table 1.
Demographic and clinic information
| Variable (category) | Total (% mean) | Year 1 (baseline) | Year 2 | Year 3 | Year 4 | p-value | Total N |
|---|---|---|---|---|---|---|---|
|
| |||||||
| Patient age n (mean) | 288 (34) | 31 (31) | 78 (34) | 101 (33) | 78 (36) | 0.290 | |
| Patient age n (median) | 288 (35) | 31 (32) | 78 (35) | 101 (35) | 78 (37) | 290 | |
| Patient age (range) | (20–65) | (20–50) | (22–63) | (22–62) | (22–65) | ||
| Unknown/Missing | 2 | ||||||
| Patient history of smoking | |||||||
| Never | 228 (79.4) | 25 (80.7) | 56 (72.7) | 83 (82.2) | 64 (82.1) | 0.712 | |
| Previous | 21 (7.3) | 2 (6.5) | 6 (7.8) | 1 (6.9) | 6 (7.7) | 290 | |
| Current | 38 (13.2) | 4 (12.9) | 15 (19.5) | 11 (10.9) | 8 (10.3) | ||
| Unknown/Missing | 3 | ||||||
| Previous history of abnormal screening results | |||||||
| No | 118 (52.7) | 7 (27.0) | 30 (49.2) | 48 (57.8) | 33 (61.1) | 0.022 | |
| Yes | 106 (47.3) | 19 (73.1) | 31 (50.8) | 35 (42.2) | 21 (38.9) | 290 | |
| Unknown/Missing | 66 | ||||||
| Percentage of families living below poverty line | |||||||
| 0.0%-25.1% | 95 (33.0) | 8 (25.8) | 33 (42.3) | 28 (27.7) | 26 (33.3) | 0.168 | |
| 25.2%-58.6% | 193 (67.0) | 23 (74.2) | 45 (57.7) | 73 (72.3) | 52 (66.7) | 290 | |
| Unknown/Missing | 2 | ||||||
| Parity | |||||||
| Null | 5 (2.0) | 3 (12.0) | 2 (2.8) | 0 (0.0) | 0 (0.0) | 0.009 | |
| 1 | 38 (14.8) | 4 (16.0) | 14 (19.7) | 10 (11.0) | 10 (14.3) | ||
| 2 | 68 (26.5) | 10 (40.0) | 16 (22.5) | 26 (28.6) | 16 (22.9) | 290 | |
| 3 | 66 (25.7) | 4 (16.0) | 21 (29.6) | 22 (24.2) | 19 (27.1) | ||
| 4 or more | 80 (31.1) | 4 (16.0) | 18 (22.4) | 33 (36.3) | 25 (35.7) | ||
| Unknown/Missing | 33 | ||||||
| Clinic site | 0.000 | ||||||
| Su Clinica | 128 (44.9) | 31 (100.0) | 39 (50.7) | 34 (34.3) | 24 (30.8) | ||
| McAllen Clinic | 157 (55.1) | 0 (0.0) | 38 (49.4) | 65 (65.7) | 54 (69.2) | 290 | |
| Unknown/Missing | 5 | ||||||
When compared across years of the program implementation, the proportion of women with a previous history of abnormal screening results, parity and clinic site differed significantly (p<0.05). The proportion of women with a history of abnormal screening was lower in Years 2–4 compared with baseline. Higher parity (>3 deliveries) was significantly higher in Years 2, 3 and 4 compared with baseline. The two clinics had a similar number of patients seen for colposcopy overall (45% (n=128) at Su Clinica vs. 55% (n=157) at the McAllen Clinic. However, these numbers differed across the years of program implementation with a higher number of women seen at the McAllen clinic compared with Su Clinica for Years 2, 3 and 4 (p= 0.000). There were no significant differences in patient age, smoking status and poverty level across the four years of program implementation.
A total of 290 women were diagnosed with CIN 2/3 with the indication for treatment with a LEEP. Of the included women with complete data available for analysis; 229 had a LEEP within 90 days of colposcopy (Figure 2). The baseline proportion of women undergoing LEEP within 90 days was 76.2%, which increased to more than 90% for Years 2–4. Additionally, the clinic ability to perform LEEP increased significantly between baseline and the follow-up years, with only 16 LEEPs performed in Year 1 (baseline) and 64, 84 and 65 LEEPs performed in Years 2–4, respectively.
Figure 2.
Proportion of women undergoing LEEP treatment within 90 days of colposcopy appointment
The results of a univariate analysis for factors potentially associated with having had a LEEP within 90 days of colposcopy are shown in Table 2. The following variables were independent predictors of undergoing LEEP within 90 days of colposcopy: Years 2 and 4 of program implementation and higher parity. Women who had colposcopy in Years 2, 3 and 4 of program implementation where more likely to receive treatment within 90 days in comparison to Year 1 (baseline). The highest impact was seen in Year 2 of program implementation (OR= 6.7, p<0.05). This increase was sustained in Years 3 and 4, however it was not statistically significant in Year 3. Having at least one child was associated with a significant increase of undergoing LEEP within 90 days of colposcopy. These differences were statistically significant in all categories of having one or more children in comparison to being nulliparous.
Table 2.
Univariate analysis of factors associated with LEEP within 90 days of colposcopy
| Variable | Univariate analysis |
||
|---|---|---|---|
| OR | 95% CI | p-value | |
|
| |||
| Year of program implementation | |||
| Year 1 (baseline) | |||
| Year 2 | 6.67 | (1.44 – 30.87) | 0.015 |
| Year 3 | 3.28 | (0.95 – 11.32) | 0.060 |
| Year 4 | 4.06 | (1.05 – 15.75) | 0.043 |
| Patient age | 1.03 | (0.97 – 1.09) | 0.239 |
| Patient history of smoking | |||
| Never | |||
| Previous | 0.35 | (0.09 – 1.35) | 0.126 |
| Current | 0.36 | (0.12 – 1.09) | 0.071 |
| Previous history of abnormal screening results | |||
| No | |||
| Yes | 0.64 | (0.23 – 1.78) | 0.397 |
| Percentage of families living below poverty line | |||
| 0.0%-25.1% | |||
| 25.2%-58.6% | 0.43 | (0.14 – 1.32) | 0.140 |
| Parity | |||
| Null | |||
| 1 | 64 | (2.83 – 146) | 0.009 |
| 2 | 37.3 | (2.59 – 537.3) | 0.008 |
| 3 | 25 | (1.84 – 339.15) | 0.016 |
| >4 | 17 | 1.38 – 209.1) | 0.027 |
| Clinic site | |||
| Su Clinica | |||
| UTMB McAllen | 0.90 | (0.35 – 2.32) | 0.825 |
The multivariate analysis results are shown in Table 3. Women who underwent colposcopy during Years 2–4 of program implementation had higher odds of receiving treatment within 90 days in comparison to those seen during Year 1 (baseline). The biggest change was seen in Year 2 of program implementation, where women were 7.48 times more likely to undergo LEEP within 90 days in comparison to those seen in Year 1 (OR = 7.48, p<0.05). These changes remained significant after adjusting for parity. Only a parity of 2 remained statistically significant after adjusting for year of program implementation. Women who were a Para 2 were 23.3 times more likely to have had a LEEP within 90 days of colposcopy in comparison to nulliparous women after adjusting for year of program implementation. The Hosmer-Lemeshow goodness of fit test results were not significant (p=0.27), we can conclude that the data fit the model well.
Table 3.
Multivariate analysis and final logistic regression model
| Variable | Multivariate analysis |
||
|---|---|---|---|
| Odds Ratio | 95% Confidence Interval | p-value | |
|
| |||
| Year of program implementation | |||
| Year 1 (baseline) | |||
| Year 2 | 7.48 | (1.35 – 41.43) | 0.021 |
| Year 3 | 4.73 | (1.04 – 21.48) | 0.045 |
| Year 4 | 5.11 | (1.00 – 25.96) | 0.049 |
| Parity | |||
| Null | |||
| 1 | 28.10 | (0.91 – 893.8) | 0.056 |
| 2 | 23.27 | (1.18 – 458.44) | 0.039 |
| 3 | 11.18 | (0.57 – 217.92) | 0.111 |
| >4 | 8.19 | (0.45 – 148.28) | 0.155 |
Figure 3 describes the mean time between colposcopy and LEEP treatment. The mean time for the overall study was 47 days (Std. dev. = 35.3, min = 0, max= 245 (8 months)). The mean number of days between colposcopy and LEEP decreased substantially between Year 1 (62 days) and subsequent years of program implementation (48, 45 and 45 days for Years 2–4 respectively).
Figure 3.
Mean time between colposcopy and LEEP treatment
Table 4 presents the univariate analysis of the variables potentially associated with the time between colposcopy and LEEP. Year of program implementation and parity had a statistically significant effect on days between colposcopy and LEEP. Women had treatment appointments 14 days earlier in Year 2 in comparison to Year 1. This period was 17 days earlier in Years 3 (p< 0.05) and 4 (p< 0.05). Women who had one or more deliveries received care more than three months earlier compared with women who were nulliparous (p<0.001). The following variables were not statistically associated with a change in days between colposcopy and treatment appointments: age, history of smoking, families living below the poverty line and clinic site.
Table 4.
Univariate analysis of variables of interest and days between Colposcopy and LEEP
| Variable | Univariate analysis |
||
|---|---|---|---|
| Coef | 95% CI | p-value | |
|
| |||
| Year of program implementation | |||
| Year 1 (baseline) | |||
| Year 2 | −14.26 | [−31.60, −3.08] | 0.107 |
| Year 3 | −17.15 | [−33.91, 0.38] | 0.045 |
| Year 4 | −17.45 | [−34.70, −0.2] | 0.047 |
| Patient age | −0.4 | [−0.90, −0.96] | 0.113 |
| Patient history of smoking | |||
| Never | |||
| Previous | 16.76 | [−0.24, 33.77] | 0.053 |
| Current | 9.13 | [−4.20, 22.47] | 0.179 |
| Previous history of abnormal screening results | |||
| No | |||
| Yes | 3.40 | [−6.98, 13.79] | 0.518 |
| Percentage of families living below poverty line | |||
| 0.0%-25.1% | |||
| 25.2%-58.6% | 1.31 | [−7.99, 10.61] | 0.782 |
| Parity | |||
| Null | |||
| 1 | −112.73 | [−151, −73.95] | 0.000 |
| 2 | −119.06 | [−151.12, −81.00] | 0.000 |
| 3 | −116.41 | [−154.55, −78.26] | 0.000 |
| >4 | −111.95 | [−149.81, −74.10] | 0.000 |
| Clinic site | |||
| Su Clinica | |||
| UTMB McAllen | −2.06 | [−10.37, 6.26] | 0.627 |
Table 5 presents the multilinear analysis of time between colposcopy and LEEP for significant covariates. Although the year of program implementation decreased the time between colposcopy and treatment appointments by 9 days (Year 2) and 13 days (Years 3 and 4) in comparison to the baseline (Year 1), these changes were not statistically significant after adjusting for parity. Having at least one delivery significantly decreased the days between colposcopy and treatment appointments (108 to 104 days, (p<0.001).
Table 5.
Multilinear analysis of covariates associated with days between colposcopy and LEEP
| Variable | Multiple linear regression |
||
|---|---|---|---|
| Coef | 95% CI | p-value | |
|
| |||
| Year of program implementation | |||
| Year 1 (baseline) | |||
| Year 2 | −9.01 | [−27.35, 9.33] | 0.334 |
| Year 3 | −12.66 | [30.57, 5.25] | 0.165 |
| Year 4 | −12.90 | [−31.33, 5.25] | 5.530 |
| Parity | |||
| Null | |||
| 1 | −104.71 | [−145.07, −64.35] | 0.000 |
| 2 | −111.93 | [−151.19, −72.66] | 0.000 |
| 3 | −108.37 | [−148.10, −68.64] | 0.000 |
| >4 | −103.71 | [−143.15, −64.27] | 0.000 |
DISCUSSION
The main finding of our study was that clinic participation in our multicomponent program increased the proportion of women who underwent LEEP within 90 days of colposcopy and cervical biopsy from a baseline rate of 76.2% to more than 90% for Years 2 to 4 of program implementation. During Years 2–4 of program implementation, the clinics successfully achieved the CDC and WHO goals of treating >90% of women with abnormal screening results and >80% of women receiving treatment within 90 days of colposcopy. The program implementation also decreased the time between colposcopy and LEEP from a mean of 62 days at baseline to 48, 45 and 45 days for Years 2, 3 and 4 respectively. These results were accomplished through community outreach, patient education and navigation, as well as provider training and support, both in person and regularly through virtual telementoring sessions.
For both outcomes, the largest difference was noted between baseline (Year 1) and Year 2, suggesting that our program had an impact early in the intervention, and this effect was sustained through Years 3 and 4 of program implementation. We hypothesize that these outcomes were due to a combination of provider training to perform LEEP according to the national clinical guidelines, training navigators to educate and navigate screen-positive women to return for diagnosis and treatment, including directly scheduling their appointments.
In our study, we noted that a previous history of abnormal screening results, clinic site and patient parity varied across the years of program implementation. We saw an increase in the proportion of patients with a history of abnormal screening results who qualified for LEEP during Years 2–3 of program implementation in comparison to baseline. We believe that this finding was due to the program navigators identifying and properly contacting an extensive list of women with a history of abnormal results who were past due for screening/treatment. This proportion went back to baseline (21%) in Year 4 of program implementation once the majority of these women had been contacted and returned for care. Parity also differed by year, with a higher proportion of women with higher parity in Years 2–4. We believe that this increase may be due to navigation and public education efforts that were held at public schools and libraries, where the participants were primarily women with children. In addition, our navigators performed inreach activities, identifying and contacting women due for cervical screening at the time of their visits in the FQHC’s Pediatric and Family Medicine clinics where their children had appointments. The proportion of colposcopy/LEEP patients differed by clinic site, however this difference was not associated with our outcomes of interest.
Parity was associated strongly with an increased odds of undergoing treatment after diagnosis in comparison to nulliparous women, as well as a reduction in time to LEEP in comparison to nulliparous women after adjusting for year of program implementation. Our hypothesis is that having more children might increase trust in the system and provide more opportunities for navigators and providers to educate women about the importance of cancer prevention. Additionally, current data for patients’ phone numbers might be more available for parous women who receive obstetric and pediatric care at the participating clinics. More research is needed in order to better understand these differences and to inform strategies for both groups of women.
Our results are similar to those described by the National Cancer Institute (NCI) patient navigation research program [33, 34] where patient navigation reduced the time to treatment initiation for breast, colorectal, prostate and cervix cancer patients. However, the patient navigation program had a significant impact on treatment initiation only after 90 days. Our study differs in that we had a more comprehensive approach which also included training medical providers in order for women to have access to appropriate treatment.
The main strength of our study is the large cohort of women who underwent screening (>14,000) and the prospectively collected, detailed data available for this analysis. This allowed us to explore potential confounding variables that could explain changes in adherence to LEEP treatment in addition to the effects of other variables such as parity.
Our study is limited in that the database did not have a direct measure of SES, however, we used county of residence as a proxy to SES as described in previous studies. [35–37] Furthermore, the study does not have a true control or pre-intervention period, with Year 1 of program implementation serving as the baseline and proxy control group. In addition, there may be other confounding variables not studied that affect time to treatment, such as availability of functioning equipment and supplies, availability of biopsy results, availability of trained providers and resources for care at the clinics, as well as the patient’s capacity to travel to care. Furthermore, we did not have detailed information about whether there were patients with CIN 2 who desired future fertility and were dispositioned to observation vs. LEEP. Given the high parity of the women in our study population, this effect was likely minimal.
In terms of the sustainability of the effects of our program, we have received additional funding to implement the same multicomponent program to other medically underserved regions of Texas and hope to provide information regarding the scalability of the program. Furthermore, existing clinics have hired patient navigators on their own, highlighting the investment FQHCs are willing to make based on the study findings.
In summary, our study showed the PRCC-Texas multicomponent program improved the proportion of women diagnosed with high-grade cervical dysplasia who underwent LEEP treatment. Our participating clinics were able to achieve the WHO target of treating over >90% of women, making the elimination of cervical cancer within reach.
Implications for Policy & Practice.
There are no clear guidelines to promote multicomponent interventions to improve treatment outcomes for patients with preinvasive disease.
Promoting interventions of multicomponent nature is relevant not only to increase cancer screening, but also to improve follow-up for diagnosis, early detection, and treatment.
Navigation may be a key factor in the success of programs and should be recommended in combination with other interventions such as medical provider training/support.
Multicomponent interventions can improve cervical cancer prevention efforts for women living in low-resource settings, where much of the burden of cervical cancer is located.
Multicomponent interventions may be a crucial factor in reaching the WHO target of treating over 90% of women with abnormal screening results.
REFERENCES
- 1.Bosch FX, et al. , The causal relation between human papillomavirus and cervical cancer. Journal of clinical pathology, 2002. 55(4): p. 244–265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bosch FX and de Sanjosé S, The epidemiology of human papillomavirus infection and cervical cancer. Disease markers, 2007. 23(4): p. 213–227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schiffman M, et al. , Human papillomavirus and cervical cancer. The lancet, 2007. 370(9590): p. 890–907. [DOI] [PubMed] [Google Scholar]
- 4.Pagliusi SR and Aguado MT, Efficacy and other milestones for human papillomavirus vaccine introduction. Vaccine, 2004. 23(5): p. 569–578. [DOI] [PubMed] [Google Scholar]
- 5.Schiffman M and Wentzensen N, Human papillomavirus infection and the multistage carcinogenesis of cervical cancer. Cancer Epidemiology and Prevention Biomarkers, 2013. 22(4): p. 553–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Castle PE, et al. , Age-appropriate use of human papillomavirus vaccines in the US. Gynecologic oncology, 2009. 114(2): p. 365–369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chatterjee A, The next generation of HPV vaccines: nonavalent vaccine V503 on the horizon. Expert review of vaccines, 2014. 13(11): p. 1279–1290. [DOI] [PubMed] [Google Scholar]
- 8.Bogani G, et al. , Assessing the long-term role of vaccination against hpv after loop electrosurgical excision procedure (Leep): A propensity-score matched comparison. Vaccines, 2020. 8(4): p. 717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Behtash N and Mehrdad N, Cervical cancer: screening and prevention. Asian Pac J Cancer Prev, 2006. 7(4): p. 683–6. [PubMed] [Google Scholar]
- 10.Safaeian M, Solomon D, and Castle PE, Cervical cancer prevention—cervical screening: science in evolution. Obstetrics and gynecology clinics of North America, 2007. 34(4): p. 739–760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Morton R, et al. , Follow‐up after treatment of high‐grade cervical dysplasia: The utility of six‐month colposcopy and cytology and routine 12‐month colposcopy. Australian and New Zealand Journal of Obstetrics and Gynaecology, 2020. 60(6): p. 959–964. [DOI] [PubMed] [Google Scholar]
- 12.Costa-Fagbemi M, et al. , Risk of cervical dysplasia after colposcopy care and risk-informed return to population-based screening: a systematic review. Journal of Obstetrics and Gynaecology Canada, 2020. 42(5): p. 607–624. [DOI] [PubMed] [Google Scholar]
- 13.Ramírez SI and Lutzkanin A, Management of Cervical Dysplasia Using Office Loop Electrosurgical Excision Procedure. Primary Care: Clinics in Office Practice, 2021. 48(4): p. 583–595. [DOI] [PubMed] [Google Scholar]
- 14.Egemen D, et al. , Risk estimates supporting the 2019 ASCCP risk-based management consensus guidelines. Journal of lower genital tract disease, 2020. 24(2): p. 132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sawaya GF, Smith-McCune K, and Kuppermann M, Cervical cancer screening: more choices in 2019. Jama, 2019. 321(20): p. 2018–2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Garner EI, Cervical cancer: disparities in screening, treatment, and survival. Cancer Epidemiology and Prevention Biomarkers, 2003. 12(3): p. 242s–247s. [PubMed] [Google Scholar]
- 17.Yoo W, et al. , Recent trends in racial and regional disparities in cervical cancer incidence and mortality in United States. PLoS One, 2017. 12(2): p. e0172548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Society., A.C. Cancer Statistics Center. 2022. [cited 2022 May 17, 2022]; Available from: http://cancerstatisticscenter.cancer.org. [Google Scholar]
- 19.Eggleston KS, et al. , Understanding barriers for adherence to follow-up care for abnormal pap tests. Journal of Women’s Health, 2007. 16(3): p. 311–330. [DOI] [PubMed] [Google Scholar]
- 20.Foundation KF State Health Facts, Health Insurance Coverage of the Total Population (CPS). 2022. [cited 2022 May 17, 2022]; Available from: https://www.kff.org/other/state-indicator/health-insurance-coverage-of-the-total-population-cps/?activeTab=map¤tTimeframe=0&selectedDistributions=uninsured&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22asc%22%7D.
- 21.Buettgens M, Blumberg LJ, and Pan CW, The uninsured in Texas: Statewide and local area views. 2018. [Google Scholar]
- 22.Ashing‐Giwa KT, et al. , The impact of cervical cancer and dysplasia: a qualitative, multiethnic study. Psycho‐Oncology: Journal of the Psychological, Social and Behavioral Dimensions of Cancer, 2004. 13(10): p. 709–728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Salcedo M, et al. , Addressing high cervical cancer rates in the Rio Grande Valley along the Texas–Mexico border: a community-based initiative focused on education, patient navigation, and medical provider training/telementoring. Perspectives in public health, 2021: p. 1757913921994610. [DOI] [PubMed] [Google Scholar]
- 24.Boom K, et al. , Perspectives on cervical cancer screening and prevention: challenges faced by providers and patients along the Texas–Mexico border. Perspectives in public health, 2019. 139(4): p. 199–205. [DOI] [PubMed] [Google Scholar]
- 25.Jones M, et al. , Identifying Barriers to Treatment for Women With Cervical Dysplasia in Rural Northern British Columbia. Journal of Obstetrics and Gynaecology Canada, 2018. 40(11): p. 1401–1408. [DOI] [PubMed] [Google Scholar]
- 26.Organization WH, Global strategy to accelerate the elimination of cervical cancer as a public health problem. 2020. [Google Scholar]
- 27.Miller JW, et al. , From cancer screening to treatment: Service delivery and referral in the National B reast and C ervical C ancer E arly D etection P rogram. Cancer, 2014. 120(S16): p. 2549–2556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.DeGroff A, et al. , When performance management works: A study of the N ational B reast and C ervical C ancer E arly D etection P rogram. Cancer, 2014. 120(S16): p. 2566–2574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Benard VB, et al. , Timeliness of cervical cancer diagnosis and initiation of treatment in the National Breast and Cervical Cancer Early Detection Program. Journal of women’s health, 2012. 21(7): p. 776–782. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bastani R, et al. , Interventions to improve follow‐up of abnormal findings in cancer screening. Cancer: Interdisciplinary International Journal of the American Cancer Society, 2004. 101(S5): p. 1188–1200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Harris PA, et al. , Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. Journal of biomedical informatics, 2009. 42(2): p. 377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.2021., S., Stata Statistical Software: Release 17. 2021.: College Station, TX: StataCorp LLC. [Google Scholar]
- 33.Freund KM, et al. , Impact of Patient Navigation on Timely Cancer Care: The Patient Navigation Research Program. JNCI: Journal of the National Cancer Institute, 2014. 106(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Freund KM, et al. , National cancer institute patient navigation research program: methods, protocol, and measures. Cancer, 2008. 113(12): p. 3391–3399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Link-Gelles R, et al. , Bias with respect to socioeconomic status: a closer look at zip code matching in a pneumococcal vaccine effectiveness study. SSM-population health, 2016. 2: p. 587–594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Taber DR, et al. , Socioeconomic differences in the association between competitive food laws and the school food environment. Journal of School Health, 2015. 85(9): p. 578–586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Feinglass J, Rydzewski N, and Yang A, The socioeconomic gradient in all-cause mortality for women with breast cancer: findings from the 1998 to 2006 National Cancer Data Base with follow-up through 2011. Annals of epidemiology, 2015. 25(8): p. 549–555. [DOI] [PubMed] [Google Scholar]