Despite its central role in cervical cancer screening, the accuracy and reproducibility of colposcopy with biopsy as a diagnostic tool are limited. Important factors that may contribute to these limitations in the United States include lack of the following: (1) standardized terminology, (2) recommendations for colposcopy-biopsy practice and procedures, and (3) quality assurance measures. Recognizing the limitations of current colposcopy approaches in the United States, the American Society for Colposcopy and Cervical Pathology (ASCCP), in collaboration with investigators from the US National Cancer Institute, set out to review evidence and develop recommendations for US colposcopy practice. This article describes the development of recommendations for colposcopy practice and procedures.
The elements of colposcopy practice in the United States are not well delineated or standardized. They likely vary depending on training of the colposcopist and volume of colposcopies performed. Although standard curricula are available, 1 the nature and quality of training vary widely. 2 This may, in part, play a role in the low sensitivity of colposcopy in the United States. 3 Studies of the individual elements of the colposcopy examination are lacking in the literature, and most elements of the examination practiced by colposcopists rely on “expert opinion” acquired during training and practice. A variety of colposcopy adjuncts have been brought to market in an effort to improve colposcopy performance, but supportive evidence has not been assessed by any US professional organization.
In response to this dearth of evidence-based consensus, the ASCCP's Board of Directors appointed the ASCCP Colposcopy Standards Committee and charged a subcommittee, working group 3, with recommending standards for the performance of colposcopy in US practices and evaluating colposcopic adjuncts. 4 These recommendations are meant to address the role and approach to colposcopy and biopsy for cervical cancer prevention.
MATERIALS AND METHODS
A systematic literature search was conducted to identify studies with relevant information about colposcopic technique, methods, instrumentation, and adjuncts. The literature search terms for these specific areas were generated at the National Cancer Institute, and reference lists were provided to the working groups. This PubMed search was performed on June 1, 2016, of English language literature between 1982 and 2015 and yielded 390 abstracts. The working group evaluated these articles for relevant results. 4
A 44-question survey was developed by the authors with input from the ASCCP Colposcopy Standards Steering Committee, the ASCCP Board of Directors, and the ASCCP's Chief Medical Research Advisor (see supplemental digital content). The survey was sent electronically to 4200 individuals who are currently or have been ASCCP members in the past 3 years. This group likely represents colposcopists with special interest in the discipline and with greater experience in colposcopic technique than the general community of women's healthcare providers performing colposcopy in the United States.
Specific elements of the colposcopic examination were evaluated, including preprocedure evaluation, colposcopic technique, number of biopsies typically performed, documentation, and follow-up management. A frequency distribution of responses was calculated. Additionally, where the number of colposcopies performed per month affected the approach to colposcopy, we stratified the survey results by volume of colposcopy practice, i.e. low-volume practice (0–5 colposcopies performed per month), moderate volume (6–20) and high-volume practice (>20 colposcopies per month). Comparisons used χ2 and trend test statistics with p values. Analyses were performed using Stata Version 14. (College Station, TX) The survey was deemed exempt from review by the institutional review board of University of Pittsburgh.
Elements of the colposcopy examination endorsed by at least 80% of survey respondents in the moderate- and high-volume groups formed the base framework for the expert consensus for the working group's recommendations. The value and validity of each element of the colposcopy examination thus identified were discussed within the working group and the larger steering committee. Elements were added or eliminated based on consensus opinion from the working group and steering committee.
Development of Recommendations
Draft recommendations were developed based on the abstracted evidence and expert consensus. The recommendations were presented to the steering committee in October 2016 and reviewed for content and consistency. Revisions were presented to all working group members for discussion and further revision in January 2017, and a vote among working group members was held shortly after. Sixty-seven percent of affirmative votes were required for approval of individual recommendations. All recommendations were approved at the first vote, and most were approved unanimously with only minor comments. After further editing and notification of stakeholder professional organizations, recommendations were posted on the ASCCP Web site for public comments between March 13 and 22, 2017, which resulted in additional modifications in response to the comments. Finally, recommendations were presented at the International Federation for Cervical Pathology and Colposcopy's 16th World Congress in Orlando, Florida, on April 5, 2017, followed by a plenary discussion. Final revisions were made by the steering committee based on comments received at this meeting. Recommendations were made for the performance of colposcopy in “minimum” and “comprehensive” colposcopy practices. Minimum refers to those basic elements needed to adequately perform colposcopy for the prevention of cancer. Colposcopists who perform a limited number of colposcopy examinations should strive to meet at least the minimum elements. The comprehensive recommendations are intended as a target for all colposcopists, but at least as an expectation for those with more extensive and consultative colposcopy practices.
Role of Funding Source
The ASCCP provided travel and logistic support for the colposcopy standards committee and working groups.
Our review of 390 articles in the English language literature retrieved from PubMed from 1982 to 2015 using search terms appropriate for colposcopic technique, methods, and instrumentation did not reveal any evidence supporting or refuting specific colposcopic methodology. 4 Our literature search, further, included colposcopic adjuncts as part of our second charge. A variety of colposcopic adjuncts have been brought to market in an effort to improve colposcopy performance. Our evaluation required that the adjuncts be currently commercially available in the United States and that evaluated studies include prospective, randomized comparisons of colposcopy and biopsy with and without use of the adjunct. We did not find sufficient evidence to recommend for or against the use of adjuncts with colposcopy.
Of the 4200 surveys distributed, we received 501 responses, representing an 11.9% response rate. Four hundred sixty-eight respondents (93.6%) indicated that they perform colposcopy. Of the 32 respondents (6.4%) not performing colposcopy, the most common reasons were practice type (pathologist, registered nurse, focus on high-resolution anoscopy), just getting started in colposcopy, not yet fully trained, and retired. Of those answering the survey, 65% were physicians, 36% were advance practice clinicians (APCs), that is, nurse practitioners, certified nurse midwives, and physician assistants, and 4% listed “others.” There was some duplication of categories resulting in more than 100% total. Respondents were mostly experienced colposcopists. The respondents were generally older (age, 78% >44 years and 36% >59 years) with more than 10 years of experience and performed a high volume of colposcopies per month. Sixty-seven percent have been performing colposcopy for more than 10 years. Thirty percent listed their practice type as “academic.” In addition, more than half of respondents reported involvement in teaching colposcopy to medical students, resident physicians, and APC students (see Table 1).
We categorized the volume of colposcopy practice by number of colposcopy examinations performed each month: 32% indicated that they perform fewer than 6 colposcopies per month (low volume); 46% perform 6 to 20 (intermediate volume); 21% perform more than 20 colposcopies per month (high volume); and 10% of respondents reported that they perform more than 50 colposcopies per month. The high-volume group was made up primarily of physicians. The proportion of physicians and APC colposcopists performing 6 or more colposcopies per month was similar (p = .254).
Respondents indicated that when preparing a patient for colposcopy, they routinely (>75% of the time) perform the following evaluations.
- Review indications for colposcopy (99%)
- Document last menstrual period (96%)
- Document contraception (95%)
- Assess pregnancy status (98%)
- Assess human papillomavirus (HPV) vaccination status (78%)
- Assess HIV status (54%)
- Document smoking status (94%)
- Obtain signed consent whether or not biopsy performed (79%)
- Obtain signed consent only if biopsy planned (24%)
The overwhelming majority of those responding (93% overall; 97% among those performing ≥6 colposcopies per month) reported use of a binocular colposcope with or without video monitor. Few (9% overall and 7% of high-volume colposcopists) used a digital imaging system or camera, and most (86% overall and 89% of high-volume colposcopists) indicated use of more than 1 magnification when doing colposcopy.
Routine technique (>75% of the time) for examining the cervix included the following elements:
- Repeat the cytology at time of colposcopy (12%)
- Examine the cervix with saline before acetic acid wash (31%)
- Examine the cervix with 3-5% acetic acid (98%)
- Examine the cervix with Lugol iodine (45%)
- Examine the cervix with a green/blue filter (80%)
- Identify the visible limits of the squamocolumnar junction (SCJ) (98%)
- Examine vulva (91%)
- Perform bimanual examination (25%)
- Perform colposcopy of vagina (55%)
- Perform digital anal examination (6%)
- Achieve hemostasis with Monsel paste or silver nitrate (99%)
The APCs and physicians included these elements in the colposcopy examination at similar rates. These technical elements were generally practiced equally among different volume practices, with the exception that fewer (26%) of those with moderate- to high-volume practices (>6 per month) examined the cervix with saline wash before applying 3% to 5% acetic acid and slightly more of this group (62%) routinely examined the vagina colposcopically.
Less than half of respondents (48%) reported routinely performing colposcopy in the face of a mucopurulent discharge. By comparison, 68% indicated that they would proceed in the presence of vaginitis and 74% after a recently treated gonorrhea or chlamydia infection. Twenty-four percent would decline to perform colposcopy in the event of any vaginal bleeding. This number increased to 84% in the presence of heavy bleeding.
The use of adjunctive technologies was uncommon. Only 8% of survey respondents indicated that they sometimes or frequently use a colposcopy adjunct: Zedscan (Zilco, Manchester, United Kingdom) <1%; Apx 100 (a prototype of Zedscan, Zilco) <1%; LuViva (Guided Therapeutics, Norcross, GA) <2%; Dysismap (Dysis Medical, Edinburgh, United Kingdom) <2%; Luma (Medispectra, Lexington, MA) <2%; and contact hysteroscopy <3%.
Generating Colposcopic Impression and Documenting Findings
Ninety percent of survey respondents routinely generated and documented a colposcopic impression. Only 25% indicated that they use the Reid (24%) 5 or Swedescore (1%) 6 scoring systems to define their colposcopic impression. There was a trend toward the use of the Reid Index in higher-volume colposcopists, ranging from 21% for low-volume colposcopists, 22% intermediate-volume colposcopists, and 36% for high-volume colposcopists (p = .022). Most respondents indicated that they document their findings using text and/or a drawing (80% and 66%, respectively). Twenty-four percent reported use of a digital imaging or print photos, some in addition to written forms of documentation. Lesion size was routinely documented by 48%, and location of lesions by 96% of respondents. Location of lesions was noted relative to a clock face by 98.5%. These practices did not differ by colposcopy volume.
There has been a rising interest in the colposcopy literature regarding the number of biopsies to obtain and the role of random biopsies in colposcopy practice. 7–10 In our study, there was wide variation in routine (>75% of the time) biopsy practice among survey respondents. Although most indicated that they would take more than 1 biopsy, the total number of biopsies and the practice of taking “random” biopsies varied among colposcopists. The decision to biopsy was often influenced by the severity of the referral cytology and colposcopy practice volume (see Table 2). With a referral cytology result of “lesser abnormalities” (atypical squamous cells of undetermined significance/HPV+, low-grade squamous intraepithelial lesion [LSIL], or negative for intraepithelial lesion or malignancy/HPV+), the worst-appearing lesion only was biopsied by 27% of respondents versus 14% for a higher-grade cytology (atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion or high-grade squamous intraepithelial lesion [HSIL]). By comparison, 72% of respondents would perform a biopsy on all lesions if the referral cytology was lesser abnormalities and 86% would perform a biopsy on all lesions for higher-grade cytology. In the setting of a referral cytology of lesser abnormalities, if the SCJ was fully visualized and no lesions were seen, 50% of respondents indicated that they would take no biopsies. On the other hand, 30% reportedly would take at least 1 random biopsy and 20% would take 2 or more random biopsies. With a cytology of ASC-H or HSIL, 20% indicated that they would take no biopsies, 26% at least 1 random biopsy, and 55% 2 or more random biopsies. Among 71% of the respondents who took random, nondirected biopsies in various situations, 39% took a single biopsy, and 31% biopsy each quadrant with no lesion.
Our survey showed variation as well in the practice of endocervical sampling (ECS). Ninety-seven percent of respondents routinely performed ECS if the SCJ was not fully visualized and if referring cytology was HSIL or ASC-H (two thirds of those who do not perform ECS indicated that a diagnostic excision procedure is routinely planned in this situation.). When ECS was performed, 21% used a curette alone, 11% used a brush alone, and 65% routinely used both curette and brush. Those with high-volume practices were twice as likely to use a brush alone as colposcopists with moderate or low volume (16% vs 8%–9%). For 22% of respondents, the patient's age influenced the decision whether to do ECS, although again high-volume colposcopists were more likely to report this than moderate- or low-volume colposcopists (32%, 17%, and 21%, respectively). Among respondents who do ECS, 36% perform ECS before biopsy and 64% perform biopsy first.
Colposcopy in Pregnancy
Practices vary with regard to performing colposcopy in pregnant women. More than half (56%) of survey respondents indicated that they would perform colposcopy in pregnancy with screening results of lesser abnormalities, i.e., HPV+ ASC-US, HPV 16/18+ NILM, or LSIL. The proportion who performs colposcopy increases to 85% when the diagnosis is HSIL, atypical glandular cells, or ASC-H. A total of 20.4% do not routinely perform biopsies in pregnancy. Twenty-six percent do biopsies only if the colposcopic impression is cancer. Thirteen percent perform a biopsy only on the single worst lesion. Fourteen percent perform a biopsy on more than 1 lesion and only 4% on all visible lesions in pregnancy. Colposcopy and biopsy practices differ by volume of colposcopy practice in pregnant women (see Table 3).
Documentation of Findings and Patient Notification
Colposcopy is a visual discipline, and the appearance of lesions revealed that after application of dilute (3%–5%), acetic acid may change with time. Findings observed by a colposcopist who lead to a directed biopsy are subjective. If dysplastic lesions are to be observed over time, treated at a different time, or evaluated by more than 1 colposcopist, meticulous documentation of findings is important. Ninety-six percent of survey respondents indicated that they document the location of lesions seen. Among those who document location, 98% identified the location by position on a clock face. Fewer than half (48%) documented the size of the lesion. Those who do are evenly divided (24%–25% each) among identifying number of quadrants, estimating size in millimeters/centimeters, and calculating the percent of cervix involved. Slightly fewer (19%) documented size using drawings, diagrams, and photos.
Follow-up with patients after colposcopy is necessary to close the loop between screening and planned treatment. More than half (53%–65%) of survey respondents indicated that they notify patients of results by phone, depending on biopsy results. Approximately one third (33%–34%) schedule an office visit to review results.
The recommendations for minimum and comprehensive colposcopy practice derived from this survey with input from the steering committee and from public comments addressed the following 6 major components of the colposcopy procedure: the precolposcopy evaluation, the examination, use of colposcopy adjuncts, documentation, biopsy sampling, and postcolposcopy procedures (see text and Table 4). Recommended elements of the precolposcopy evaluation for minimum colposcopy practice include indications for colposcopy, pregnancy status, menopausal status, history of hysterectomy, and obtaining informed consent. Expectations for a comprehensive colposcopy examination include, in addition to the elements recommended for minimum practice, history relative to cervical cytology, colposcopy and treatment, parity, contraception, smoking history, HIV status, and HPV vaccination status. For minimum colposcopy practice, the colposcopist should examine the vulva and vagina grossly and examine the cervix with magnification after application of 3% to 5% acetic acid. In addition, a comprehensive examination includes evaluation of the upper vagina with magnification, examination of the cervix after application of 3%–5% acetic acid using multiple magnifications, and the use of both white light and a red-free (blue or green) filter. Documentation in minimum colposcopy practice should be at least in text format and, in the comprehensive practice, should include a diagram or photograph, annotated if possible with findings imported into the electronic medical record. The visibility or nonvisibility of the SCJ, presence or absence of acetowhitening, and colposcopic lesions should be documented for all colposcopies. Each colposcopy should include a colposcopic impression (benign–normal/LSIL/HSIL/cancer). In addition, the colposcopist performing a comprehensive examination should document whether cervical manipulation is needed to completely visualize the SCJ, and if lesions are present, the extent of lesion visualized (fully/not fully), lesion size and location, and description (color, contour, border, vascular changes). In both minimum and comprehensive practice, biopsies, if indicated, should be taken at the SCJ and ECS, if performed, should be documented. In comprehensive practice, the location of any biopsies and the method of ECS, if done, should be documented. Finally, arrangements should be made to notify the patient of the biopsy and ECS results with follow-up management plans for all colposcopies. For a comprehensive practice, documentation of these arrangements should be included. As for our second charge, based on our review of the literature, we were unable to recommend for or against the use of colposcopic adjuncts.
Without level 1 or 2 evidence to support performing most elements of the colposcopic examination, we performed a survey of current or recent ASCCP members to help define “expert practice” in the United States. The results of our survey demonstrate that colposcopists show a surprising uniformity of practice. The following certain elements have a high level of consistency: use of a binocular colposcope at several magnifications, application of dilute acetic acid, evaluation using a blue/green filter, assessing completeness of visualization of the SCJ, the nature and location of lesions, and documenting the findings. There are differences in practice regarding other elements of the procedure, including routine use of Lugol iodine, use of saline wash, routine colposcopy of the vagina, and documenting the size of lesions. There are also variations regarding colposcopy during pregnancy, performance of colposcopy for minimally abnormal screening results, number of biopsies to take, the role of the random biopsy, and routine ECS. These differences vary with volume of colposcopy practice. Although most colposcopists surveyed, regardless of the volume of their practice, take biopsies of all lesions seen, those with the highest volume practice (>20 colposcopies per month) were more likely than less active colposcopists to take no biopsies if the SCJ is fully visualized and no lesions are seen, regardless of the antecedent cytology. When lesions are identified, these “high-volume” colposcopists were also more likely to biopsy only the worst-appearing lesion. There is controversy in the literature over the role of taking random (nondirected) biopsies. 8–10 This was addressed by Working Group 2 of the Colposcopy Standards Committee and they found no added benefit for women with low risk of precancer. 11 The evidence from the literature is clear that sensitivity is increased with taking multiple biopsies. 7,10 Working group 2 underscored this in most situations using a risk-based approach. 11
Colposcopic examination with biopsies for pathologic assessment is the criterion standard in the diagnostic evaluation of cervical dysplasia. Previous studies of colposcopy report a sensitivity of only 53.6% to 69.9%. 3,7 Furthermore, studies have demonstrated poor interobserver and intraobserver reliability. 12 Attempts, therefore, have been made to improve the diagnostic properties of colposcopy by developing new adjunctive technologies that exploit the altered biochemical and morphological changes present in dysplastic cervical cells. The following 3 main techniques are currently in development: fluroscence and reflectance spectroscopy, dynamic spectral imaging, and optical coherence tomography. Only one of these, dynamic spectral imaging, is Food and Drug Administration approved and marketed in the United States. Among colposcopists taking our survey, only 5% indicated that they sometimes or frequently used one of these types of adjuncts.
The results of our survey should not be generalized to all US colposcopists. The survey participants were current or recent ASCCP members. The results are necessarily biased toward the practice preferences of clinicians with an interest and/or clinical focus on colposcopy and prevention of female lower genital tract cancer.
In developing the ASCCP Colposcopy Standards for colposcopy procedures, we based the framework for our recommendations on practice patterns endorsed by at least 80% of those survey respondents in the combined moderate- and high-volume groups. Interim recommendations were then developed with input by the entire ASCCP Colposcopy Standards Committee, Board of Directors, and Chief Medical Research Advisor. They were modified again after considering comments on a national electronic bulletin board posted on the ASCCP Web site (http://www.asccp.org) and after comments solicited during presentation at the International Federation of Cervical Pathology and Colposcopy World Congress April 5, 2017. Recommendations for both minimum and comprehensive colposcopy practice were established (see Table 4). The minimum colposcopy practice should be the baseline requirements for a colposcopy examination in the United States. This will be acceptable for those who perform colposcopies infrequently. Most colposcopists, however, should be able to practice at the comprehensive level.