Test results and follow-up care stemming from an ED-based cervical cancer intervention

Abstract

Background:

More than 4300 lives are lost to cervical cancer in the US annually. The survival rate is nearly 100% when precancerous lesions are identified and acted upon appropriately. A randomized controlled trial of two Emergency Department-based (ED) cervical cancer screening interventions demonstrated that patients in each condition received screening at rates far above an untreated historical control group. While screening represents a necessary first step in preventing cervical cancer morbidity and mortality, a lack of data on subsequent clinical care limits appreciation of ED-based screening efforts.

Objective:

The current study describes test results and subsequent care observed among patients who received documented cervical cancer screening following an ED-based intervention.

Methods:

Information on HPV infections, Pap test results, and cervical cancer diagnoses documented in the electronic health record were collected from women ages 21–65 who were (a) enrolled into the above-mentioned trial and (b) received cervical cancer screening over the 365-days post enrollment.

Results:

181 women received screening. Of those who had a Pap test, 12% had abnormal findings. Compared to a national rate of 3.8%, significantly elevated risk was observed (Z = 5.55, p < 0.001). High-risk HPV infection was observed in 15% among those tested, elevated from an estimated nationwide prevalence of 11% (Z = 1.63, p = 0.052). Notably, nearly half of patients who went on to have colposcopy and cervical biopsy did not have a routine women’s health provider.

Conclusion:

ED-based cervical cancer screening interventions have meaningful positive impact on subsequent testing and care.

1. Introduction

Over 13,000 new cases of cervical cancer are diagnosed each year in the United States and more than 4300 lives are lost [1]. Fortunately, cervical cancer is one of the most preventable forms of cancer. Timely cervical cancer screening among individuals 21–65 years of age is an effective method for reducing/eliminating cervical cancer related mortality and morbidity. The United States Preventive Services Task Force (USPSTF) recommends screening for cervical cancer using cytology (i.e. Pap test) every 3 years for women age 21 to 29 years and screening using HPV testing every 5 years for women age 30–65 (or every 3 years using cytology alone). Screening for human papilloma virus (HPV) seeks to identify the vaginal presence of high-risk HPV genotypes (hrHPV) that cause cervical cancer (e.g. HPV 16 and HPV 18 account for 70% of cases worldwide), whereas the Papanicolaou test examines cervical epithelial cells for the presence of precancerous lesions. Presence of hrHPV infection or a positive Pap test, when evaluated in the context of patient medical history, can lead to colposcopy (i.e., magnified view of the cervix using a colposcope) with cervical biopsy.

Timely screening currently prevents roughly 70% of cervical cancer deaths [2], though estimates imply that rate would increase to 83% with universal adherence to screening guidelines. The survival rate is nearly 100% when precancerous lesions are identified and acted upon appropriately. Unfortunately, many eligible patients do not stay up-to-date with recommended cervical cancer screening.

A variety of efforts have been conducted aimed at increasing screening rates [3–5]. We conducted a randomized controlled trial comparing screening behaviors of Emergency Department (ED) patients 150 days post-visit between individuals who received a referral for screening to those who received a referral and a series of six text-messages grounded in the Theory of Planned Behavior [6,7] and Self-Determination Theory [8,9]. Results showed that patients in each condition received screening at rates far above an untreated historical control group, though differences between conditions were limited [10]. In much of the cervical cancer screening literature downstream clinical outcomes are not addressed. While screening represents a necessary first step in preventing cervical cancer morbidity and mortality, the lack of data on subsequent clinical care can limit appreciation of screening efforts in the ED and other opportunistic settings. The current study addresses this gap by describing test results and subsequent care observed among patients who received documented cervical cancer screening following an ED-based intervention.

2. Methods

2.1. Patient sample

Patients included women, age 21–65, who were (a) enrolled into the above-mentioned ED-based cervical cancer screening clinical trial and (b) received cervical cancer screening documented in their electronic health record (EHR) over the 365-days post study enrollment. Only patients who spoke English, Spanish, or American Sign Language were included in the trial.

2.2. Methods

Representatives of the Emergency Department Research Associates program (EDRA [11–13]) completed chart abstraction for each patient. EDRAs reviewed charts from the UR Medicine system, as well as from Rochester Regional Health Information Organization to provide evidence of screening results and clinical care at sites outside the home system. Information on hrHPV infections (HPV 16, HPV18, and other hrHPV), Pap test results, and cervical cancer diagnoses was collected. Data were entered directly into REDCap, a secure online data capture application that is HIPAA compliant [14]. The Research Subjects Review Board at University of Rochester approved all study activities, and participants provided informed consent for review of their charts.

2.3. Plan of analysis

Rates of screening results were described using frequencies and percentages. We then related follow-up testing rates among those with abnormal screening results using conditional probabilities.

3. Results

3.1. Sample description

A total of 181 participants from our clinical trial [10] received cervical cancer screening documented in the EHR during the 365 days following identification of screening needs in the ED (96 referral and text-message intervention/85 referral only intervention; 17% of the randomized cohort). Results indicated that, for every 6 patients who received one of our ED-based interventions, we observed a documented cervical cancer screening in the next year (181/1089). This number is quite favorable relative to historical controls, where 1 in 10 ED patients non-adherent with screening guidelines subsequently receive screening without intervention [10].

The mean age was 35.8 years (SD = 12.5). Demographic characteristics of this sample are presented in Table 1.

Table 1.

Sample demographics.

	Frequency	%
Race
White	122	67%
Black/African American	43	24%
Asian	9	5%
Multi-racial	5	3%
Other	2	1%
Hispanic Ethnicity	33	18%
Educational Level
> High School Degree	10	6%
High School Degree or Equivalent	55	30%
Some College	63	35%
4-year College Degree	30	17%
Professional Degree (MA, PhD, MD)	21	12%
Insurance Coverage
Private Insurance	93	51%
Medicaid	66	37%
Medicare	14	8%
No Insurance	4	2%
Has a Routine Provider for Women’s Health Issues	113	62%

3.2. Screening results

Results of cervical cancer screenings are presented in Table 2. A total of 28 participants had an abnormal result from a Pap test or hrHPV test (16%); 21 had abnormal Pap results (12% of those with a Pap test) and 16 had an hrHPV infection (15% of those with an hrHPV test; 9 had both an abnormal Pap test and an hrHPV infection). Results indicated that the number of ED patients needed to intervene upon to see an abnormal cervical cancer screen was 33 (28 positive cases/1089 total cases).

Table 2.

Cervical cancer screening results.

	Frequency	%
Pap Test and/or hrHPV Test Results
Negative/No Abnormality	133	74%
Abnormal	28	16%
Cannot be determined	20	11%
PapTestOutcomes (n = 179)
Negative/No Abnormality	139	78%
Atypical Squamous Cell ofUndetermined Significance (ASCUS)	9	5%
Low-Grade Squamous Intraepithelial Lesion (LSIL)	9	5%
Abnormal Results - Unspecified	3	2%
Unknown – Results Unavailable	13	8%
Inadequate Sample	6	3%
HPV Test Outcomes (n = 109)
Negative for all hrHPV	90	83%
Positive for HPV 16 and/or 18	3	3%
Positive for other hrHPV	13	12%
Unknown – Results Unavailable	3	3%

The national rate of abnormal findings on Pap tests is 3.8% [15]; the observed rate of 12% indicates significantly elevated risk in this ED patient population (Z = 5.55, p < 0.001). A large study of racial-diverse patients in the US estimated hrHPV infection at 11% [16]; the observed rate of 15% also indicates elevated risk (Z = 1.63, p = 0.052).

3.3. Follow-up testing

Among those 28 patients with abnormal Pap test or hrHPV results, 7 received colposcopy (25%), 6 received cervical biopsy (21%), and 1 patient was diagnosed with cervical cancer. Of note, 3/7 colposcopies and 3/6 biopsies were received by patients who did not have a routine provider for women’s health concerns.

4. Discussion

Cervical cancer screening identifies hrHPV infections and/or precancerous cervical lesions at a point where medical intervention, if indicated, has a high likelihood of success [17]. ED-based interventions can increase screening rates [10,18] within this population. The results of the current study underscore the potential importance of targeting patients in this setting, as the rates of abnormal Pap tests and hrHPV infections observed in this sample of the ED patient population are significantly elevated relative to overall national estimates.

Despite a lack of pre-existing patient-provider relationship, our interventions through the ED facilitated increased screening and identification of screening results that required additional testing or treatment. Furthermore, among those patients with abnormal results upon initial screening, we found that follow-up testing occurred at very similar rates among those with a women’s health provider and those without, supporting the use of screening interventions through the ED that seek to link patients with community resources. Results indicated that for every 6 patients we intervene upon, we observe one documented cervical cancer screening in the EHR. Furthermore, for every 33 patients, we observed at least one abnormal cervical screening test. These estimates, while encouraging in their own right, are likely downwardly biased due to screening that occurred outside the UR Medicine medical systems (e.g., Planned Parenthood). In general, the current study highlights that brief, inexpensive interventions require few patient approaches to obtaining meaningful clinical data.

4.1. Limitations

There are several limitations to the current work. First, as mentioned above, a minority of cases had results we could not see because they were outside of our system. Any abnormal results in these tests would further strengthen the observed findings. Second, the study was conducted in a single healthcare system, such that replication in heterogeneous settings would validate or refine observed estimates. Finally, the method employed limited evaluation of shared decision-making discussions and provider recommendations for subsequent screening intervals and modes.

4.2. Conclusion

ED-based cervical cancer screening interventions have meaningful positive impact on subsequent testing and care.