Postcoital bleeding (PCB) is common, especially in women using hormonal contraception, and it is frequently associated with a high degree of anxiety. The majority of the women with PCB have no underlying cervical pathology; however, many of these women are referred to colposcopy clinics. Such visits use valuable specialized colposcopy resources and can lead to unnecessary interventions.

In this study we evaluated patients referred to the colposcopy clinic with postcoital bleeding and focused on those who had a history of a negative smear within the prior three years. The aim of the study was to establish the prevalence of high grade CIN and cervical cancer in this population.

We conducted a retrospective study of all patients referred to our colposcopy clinics with PCB during the 69-month period from March 2005 to December 2010. Patients were identified from the colposcopy database. Smears were reviewed from the three years prior to their first visit to the colposcopy unit and included the smears performed at those visits. Patients who had negative smears were followed up for a minimum of one year to calculate the prevalence of high grade CIN & cervical cancer in this group of patients. Patients who had negative smears but suspicious looking cervices were excluded. We reported the highest grade of cervical neoplasia identified in any biopsies taken during the follow up period. The majority of cervical biopsies were, however, taken at the time of first visit. Where the smear history was not available using the usual results server lookup, the full smear history report was requested from the West Yorkshire Cervical Screening Authority for patients in whom a biopsy had shown CIN2 or greater.

During the study period there were a total of 1470 new referrals for colposcopy for PCB. The overall prevalence of CIN was 12.1% (179/1470) and of high grade CIN was 3.8% (56/1470). There were six cases of cervical cancer (0.4%) (6/1470) (Table 1). In addition, there was one case of CGIN and one case of endometrial cancer. All the cases of cervical cancer had abnormal smears at the time of referral (five had severe dyskaryosis and one had suspicion of invasion). Of those 6 patients, only one had regular cervical smear tests and all her previous smears were negative. One patient never had smear before and her first smear was in the colposcopy clinic when she was 28 years old. The other 4 patients missed their routine recall smears with the previous smear being 5 to 13 years prior on average.

Table 1: Patients referred to colposcopy clinics found to have high grade CIN and/or cervical cancer.

1074 out of 1470 women had a negative smear within the three years prior to referral (Table 2). The colposcopic findings in this group were: normal cervix in 35.5% (381/1074), ectropion in 25.2% (271/1074), and HPV changes in 14.4% (155/1074) (Table 3). A total of 579 biopsies were carried out, from which 20 patients had CIN2 (1.9%) (20/1074) and four had CIN3 (0.37%) (4/1074) (Table 4). One patient had CGIN (0.09%) (1/1074) in the initial cervical biopsy but not in a subsequent LLETZ. Two patients had endometriosis reported in the cervical biopsy. The prevalence of CIN was 9.0% (97/1074) and of high grade CIN (CIN2/3) was 2.2% (24/1074). There were no cases of cervical cancer detected in this group.

Table 2: Smear results in patients with PCB at the time of referral to the colposcopy clinic.

Table 3: Colposcopic findings in patients with PCB and prior negative smear.

Table 4: Biopsy findings in patients with PCB and negative smear.

Patients referred with negative cytology had a lower prevalence of histological abnormality than those with no cytology. 319 patients out of 1470 women had no smear done or their smear results were not available. The prevalence of CIN in this group of patients was higher than the smear-known group at 14.7% (47/319, p<0.01, chi-square test) but that of high grade CIN was not significantly different at 3.1% (10/319, p=0.36) (Table 5). There was one case of endometrial cancer that was confined to an endocervical polyp but no cases of cervical cancer were detected in this group.

Table 5: Biopsy findings in patients with PCB with no smear or results of smear not available.

PCB is a common problem and can be a major cause of anxiety in women given the reported association with cervical pathology. However the literature shows a variation in the prevalence of cervical cancer & CIN among women with PCB.

Anorlu et al. [1] report a series of 885 women in a clinic in Lagos Nigeria 1998-1999. Dyskaryosis was significantly higher in symptomatic compared to asymptomatic cases (6.1% vs. 3.4% p<0.01) with 9.3% of patients with postcoital bleeding demonstrating dyskaryosis.

A retrospective study by Rosenthal et al. [2] of 314 women with PCB showed 12 cases of invasive cancer (3.8%): ten were cervical or vaginal cancers and two endometrial cancers. Eight out of the ten cervical /vaginal cancers were clinically apparent. Four of these ten had normal smears before being referred for further investigation of PCB. Two of those were visible only with the aid of the colposcope. Thus 0.6% (2/314) of women attending the gynaecology service with PCB, and a normal looking cervix and normal smear had invasive cancer of the cervix. Cervical intraepithelial neoplasia was found in 54 women (17.2%).

In a retrospective study of 284 women with PCB, 166 women had either no or normal smear records (group 1) and 118 had abnormal smears (group 2) [3]. In group one, 72 women were more than 35 years old and 94 were 35 or younger. The rate of cervical cancer was 3.6% in group 1 (1.2% when excluding those with no smear record) and 5% in group 2. The equivalent figures for CIN were 9% and 66.1% respectively. There was no significant difference in the prevalence of cervical cancer or CIN between women older or younger than 35 years in group one. Neither age nor duration of PCB was shown to be a reliable indicator for cervical cancer in this study. However, another study of 137 women presented with post-coital bleeding showed that significant cervical pathology was found in 28 women (20.4%); 14 (50%) were <35 years of age, 26 (92.8%) had PCB for >4 weeks, whereas seven (25%) suffered severe episodes, suggesting that the duration, but not age or severity is important [4].

Another retrospective study of 142 women with postcoital bleeding reported a total of 27 (19%) had cervical intraepithelial neoplasia (CIN) out of which there were 15 (10.6%) cases of high-grade disease (CIN2 and CIN3) [5]. Out of the 102 women who had a negative smear within the three years prior to referral, 20 had CIN (19.6%) and ten of those were high grade CIN (eight with CIN 2 and two with CIN3) (9.8%). In this cohort there were no cases of invasive cancer of lower genital tract. Similarly, there was no case of invasive cancer of the lower genital tract cancer detected in another retrospective study of 248 women referred with PCB [6]. The prevalence of CIN in that study was lower than the other studies reported at 6.8% (12/248).

Our study, in an unselected population of women presenting with postcoital bleeding as the primary complaint, confirms that the majority of women with a normal-appearing cervix and a prior negative smear test have no underlying significant cervical pathology. Although the overall prevalence of CIN was 9% we found high grade CIN in only 2.2% of this group of patients and mostly this was CIN2. The prevalence of invasive cancer was zero. A similar retrospective study found high grade CIN in 1.6% (1/64) of patients referred to colposcopy with postcoital bleeding and negative smears, in which no cases of cancer were detected [7]. Similarly, another retrospective study of 87 women with postcoital bleeding and negative smears found CIN in 6.9% & there were no cases of invasive cancer [8].

There are wide variations in the management of women with postcoital bleeding in the UK [3,9]. The available guidelines recommend referral to colposcopy clinics in the presence of an abnormal smear, suspicious cervix, or persistent and/or repeated bleeding. NICE guidelines state that women with postcoital bleeding should have full pelvic examination, including speculum examination, by the primary health care professional and those patients who have clinical features suspicious of cervical cancer should be referred urgently [10]. A cervical smear test is not required before referral, and a previous negative smear result is not a reason to delay referral.

Similarly, the National Health Service cervical screening programme recommends that women presenting with symptoms of cervical cancer– such as postcoital bleeding (particularly in women over 40 years) – should be referred for gynaecological examination and onward referral for colposcopy if cancer is suspected [11]. Examination should be performed by a gynaecologist experienced in the management of cervical disease (such as a cancer lead gynaecologist). The equivalent Scottish guidelines recommend urgent referral of women >35 years with persistent (>4 weeks) postcoital bleeding and early referral for women with repeated unexplained postcoital bleeding [12].

It is clear from the literature that there is variation in the reported prevalence of underlying CIN and cervical cancer in women presenting with postcoital bleeding, with or without normal smears. This could be partially attributed to different populations being studied where risk factors for CIN or cervical cancer vary. However, in most of the studies the sample size is small, less than 400, and further studies of larger volume were needed to provide a reliable estimate of CIN and cervical cancer prevalence. Furthermore, the effect of the implementation of HPV testing in the management of this group of patients needs to be assessed.

Until more evidence is available, we have demonstrated that women presenting with postcoital bleeding and an up-to-date smear history do not need to be seen primarily in the colposcopy clinic and can safely be referred to a general gynecology clinic. Referral of those patients to the colposcopy clinic may be considered when the cervix cannot be visualized satisfactorily, a suspicious lesion is present on the cervix, or the patient has an abnormal smear history. Colposcopy should not be considered a primary procedure of investigation for young patients presented with postcoital bleeding who have negative smear history and a normal looking cervix.

Women with postcoital bleeding overall seem to have a higher incidence of CIN than the general population. However, postcoital bleeding in a woman with a negative prior smear and a normalappearing cervix is rarely associated with significant pathology. Such patients should not be referred directly to the colposcopy clinic. Referral indications for colposcopy in this group include unsatisfactory visualization of the cervix or associated lesions suspicious for cancer. There is no consensus for management of these patients and further study is needed to standardize the management of post-coital bleeding.

All authors have no conflicts of interest to declare.

RO: substantial contributions to the conception and design of the study, acquisition of data, analysis and interpretation of data, and drafting the article. SS: substantial contributions to the conception and design of the study, analysis and interpretation of data, and final approval of the version to be published.

No approval was needed to perform the study.