METHODS

Study population

Consecutive patients presenting with a new problem at the GUM clinic in Coventry and having a blood test for routine syphilis screening were asked to participate voluntarily. Study clinics represented equally the different daily clinical sessions. Participants were not known to be HIV positive. After verbal consent, eligible patients were asked to complete a self administered questionnaire. During the medical consultation, demographic details and sexual history were taken. At the end of the consultation, patients were counselled (pretest counselling) for 5–10 minutes about genital herpes and its serotesting and were asked to complete another copy of the same questionnaire. When the results were given a week later, patients were counselled again (post-test counselling) and asked whether they regretted having the test. Post-test counselling (3–5 minutes) discussed the results of serotesting and any other points raised by patients, and answered their questions.

The questionnaire (table 1) can arbitrarily be divided into three sections (A, B, C): section A (questions 1–8) to assess patients’ knowledge about genital herpes, section B (questions 9, 10, 11) to evaluate patients’ feeling towards genital herpes infection, and section C (questions 12, 13) to examine patients’ attitude towards serotesting. We evaluated the answers to section A (Q1–Q8) in two different ways. Firstly, we applied a scoring system in which we marked the answer to each question for each patient, before and after counselling, giving +1 for the correct answer, −1 for the incorrect answer, and 0 for “not sure.” Secondly, we calculated the percentage of correct, incorrect, and “not sure” answers for all eight questions of all the participants, before and after counselling.

Pretest counselling included all the points mentioned in the questionnaire. Counselling was carried out by the same physician to maintain consistency. Study investigations were carried out during routine busy clinics. Before discharge, patients were advised to contact the study physicians or the health advisers should they have any concern. Ethics committee approval was obtained for the study.

Laboratory methods

All samples were tested in the local PHLS laboratory. Collected sera were stored at −20°C until the time of processing. A commercially available enzyme immunoassay (EIA) was used. This was based on a type specific glycoprotein G1 (gG-1) from HSV-1 and gG-2 from HSV-2 (Gull/Meridian gG EIA, Meridian Diagnostics Inc, Salt Lake City, UT, USA). The laboratory methods have been described elsewhere.^10–12

Statistical analysis

Logistic regression models were used to identify the potential risk factors for HSV-1 and HSV-2 infections. Potential risk factors for HSV-1 and HSV-2 antibodies were identified using univariate logistic regression models. These factors were then used in the multivariate logistic regression model. The χ² test was used to identify any difference between the sexes in the prevalence of HSV-1 and HSV-2 antibodies. The difference in age between males and females was tested using the Mann-Whitney U test.

Change in score from before counselling to after counselling for questions 1–8 of the self administered questionnaire was analysed using the Wilcoxon signed rank test for paired data. In question 9, we compared those who considered genital herpes is worse than chlamydia or gonorrhoea against those who considered that chlamydia or gonorrhoea is worse than genital herpes. The change in the opinions before and after counselling was analysed using the McNemar’s test for paired data. The Wilcoxon signed rank test for paired data was also used to test the difference between the distribution of scores to questions 10 and 11. For questions 12 and 13, where the answers were given in three categories, the difference was analysed using the χ² test. We acknowledge that although the χ² test does not incorporate the paired structure of the data, it was used for simplicity.

RESULTS

Two hundred and twenty three (107 males and 116 females) of 230 (97%) eligible clinic attenders, over a 6 month period, completed the questionnaire. Four of those who did not participate said they did not think they needed it, and three did not give a reason. All questionnaires were suitable for analysis, although data were missing in a few of the questions. Sufficient serum for testing was available in 220 participants. On testing, three samples gave equivocal results for both types of HSV, two for HSV-1 and one for HSV-2. Blood testing was not repeated for equivocal results. Table 2 shows participants’ general characteristics. The age was significantly different between males and females (p<0.001). Results of serotesting for HSV-1 and HSV-2 antibodies are shown in table 2.

Only 14/42 (33%) participants with antibody to HSV-2 gave a history of genital herpes. In contrast, 14/29 (48%) patients giving a history of genital herpes had HSV-2 antibody. Four per cent of patients (3/68) with no antibodies to either HSV-1 or HSV-2 had a history of genital herpes. The prevalence of HSV-2 antibody was similar in patients with or without antibody to HSV-1 (25/128 (20%) and 18/86 (21%), respectively). A previous history of genital herpes was reported by 7/18 (39%) patients with antibody to HSV-2 only, compared with 7/24 (29%) patients with both HSV-1 and HSV-2 antibodies.

Tables 3 and 4 show the relation between antibodies to HSV-2 and HSV-1 and some of the studied sociodemographic, sexual behaviour, and clinical factors. Variables of sexual orientation, injecting drugs, and sex with prostitutes were not tested, because of the small number of patients involved. Unemployed people and housewives were not included in the tables in the variable of occupation because of their heterogeneity. For HSV-2 antibody, increased risk of infection was associated with increasing age, black race, lower education, earlier age of sexual intercourse, increasing years of sexual activity, number of lifetime partners (⩾5), not using condoms, other STIs, patient’s history of genital herpes and manual jobs. However, in the univariate analysis, only being black or having a history of genital herpes had a significant association with the HSV-2 antibody. This was confirmed in the multivariate analysis. A rise in the risk of HSV-1 infection was observed with increasing age, being black, lower education, increasing years of sexual activity, number of lifetime partners (⩾5), patient history of cold sores, other STIs, and manual jobs. In the univariate analysis, the association was significant only for age (⩾35 years), lower education, increasing years of sexual activity (between 14 and 25 years), and patient history of cold sores. However, this association was independently significant only for lower education, increasing years of sexual activity (between 14 and 25 years), and patient’s history of cold sores.

The answers to the questionnaire are given in table 5. Answers were compared before and after counselling. For section A (Q1–Q8), the total scoring is presented in the table. Overall, 184/217 (85%) participants showed improved scores after counselling compared with scores before counselling. One hundred and six out of 217 participants (49%) achieved ⩾50% improvement. Of those who did not score the full mark before counselling, 114/199 (57%) achieved it after counselling. Eleven out of 217 participants (5%) scored less after counselling, while 14/217 (6%) scored the full mark before and after counselling. Of the answers to section A (Q1–Q8) by all participants before counselling, 59% were correct, 12% incorrect, and 28% “not sure.” After counselling, these changed to 92% correct, 5% incorrect and 3% “not sure.”

The majority of participants were aware that genital herpes is an STD (184/222 (83%) before and 204/218 (93%) after counselling) and that it is caused by a virus (147/222 (66%), increased to 211/218 (97%)). Only 122/222 (55%) were aware that the virus causing cold sores can cause genital herpes (increased to 206/218, 94% after counselling). Fifty four per cent of the answers to the questions about unrecognised infection and asymptomatic shedding were correct (increased to 90% after counselling). Before counselling, about two thirds of participants expected the effect on them if they (Q10) or their partners (Q11) had been diagnosed with genital herpes infection to be moderate to severe (152/222 (68%) and 135/222 (61%), respectively). One fifth of participants expected the effect, in both situations, to be very severe (scale on table 1). This opinion was not significantly changed after counselling (table 5). Only three patients regretted having the test. Four patients contacted the department within six months of receiving the results for more counselling. All had seropositive HSV-2 antibody and three had no previous history of genital herpes. Their contact was primarily for more discussion and understanding.

DISCUSSION

To our knowledge, this is the first study to assess the seroprevalence of both HSV-1 and HSV-2 antibodies among GUM clinic attenders in a UK metropolitan city. It is also the first in the United Kingdom to assess patients’ knowledge and attitude before and after counselling, and to evaluate their perception of results.

Two serotypes of HSV are recognised, with antibodies usually detectable 8 weeks after the onset of infection.¹³ HSV-2 is almost always transmitted sexually,¹ and detection of HSV-2 antibody almost always indicates previous genital herpes infection. Understanding of the seroepidemiology and sociology of genital herpes cannot be obtained without accompanying information regarding HSV-1 infection patterns.¹ HSV-1 is common in the general population and is often acquired non-sexually in childhood, causing orolabial infection. HSV-1, however, can cause genital herpes, and the proportion caused by HSV-1 is increasing. In the United Kingdom, the incidence of new cases of genital herpes caused by HSV-1 can be as high as 50%.¹⁴ However, HSV-2 reactivation and recurrence remain the more prevalent.¹⁵ Thus, although the detection of HSV-2 antibody is essentially indicative of genital herpes infection, its absence does not exclude it. Detection of HSV-1 antibody on the other hand cannot indicate whether the site of infection is oral or genital. Our study population was informed about the difficulties in this area.

The seroprevalence of genital herpes varies worldwide. A seroprevalence rate of HSV-2 antibody as high as 60–90% has been reported in several developing countries. In developed countries, it is estimated that as many as 20% of the general population may be seropositive for the HSV-2 antibody. The prevalence of genital herpes is often substantially higher in high risk populations—for example, STD clinic attenders.^1,16–20 The difference between studies may reflect the difference in ethnicity, social background, and sexual lifestyle between different populations. The rate of HSV-2 seropositivity in our study (overall 20%, males 18%, females 21%) is similar to the only study undertaken in a STD clinic in London²¹ (overall 22.7%, males 17.3%, females 24.5%), but higher than in a study done in an STD clinic in a district hospital outside London²² (overall 15.6%, males 9.9%, females 18.7). The seroprevalence rate of HSV-2 antibody among blood donors in London was reported as 3% in males and 12% in females,²¹ while outside London it was 3.2% and 7.8% respectively.²³ In most studies, the rate is higher among females than males. In our study, it is higher in females although the difference is not significant. This may be explained, at least in part, by the older age, the greater number of STDs, the greater number of partners, and the higher proportion of black people among the males (table 2).

In most studies, HSV-2 seropositivity is independently associated with increasing age, lower income, lower education, increased years of sexual activity, previous STDs, multiple sexual partners, and earlier age of first intercourse.^{1,16,19–21,24} In our study, only being black or having a history of genital herpes had a significant association with the frequency of HSV-2 infection in the multivariate analysis. An increase in the number of participants in the study might have revealed more significant factors. HSV-2 seropositivity has been suggested as an objective marker of sexual behaviour.^1,21

Risk factors for the acquisition of HSV-1 usually include increasing age, lower socioeconomic status, and black ethnicity.^1,24 Rates of 60–90% in Europe and rates higher than 90% in Africa have been reported.^1,24 A study undertaken in an STD clinic in London reported the seroprevalence of HSV-1 as 59.5%.²⁵ In our study, a similar prevalence (60%) was found with no difference between males and females. The prevalence was independently related to lower education, increased years of sexual activity (15–24), and patients’ history of cold sores.

In our study, only a third of the study participants with HSV-2 antibody gave a history of genital herpes. This is consistent with other studies and highlighted the poor sensitivity in the clinical diagnosis of genital herpes.^1,16,19–21 Although we do not know which type occurred first, patients with antibodies to HSV-1 and HSV-2 (that is, co-infection) had less history of genital herpes (symptoms/diagnosis) than those with antibody only to HSV-2 (29% compared with 39%). On the other hand, prevalence of HSV-2 antibody is similar in patients with negative or positive antibodies to HSV-1 (21% and 20%, respectively). This is consistent with most studies and indicates that although previous infection with HSV-1 may not decrease the incidence of infection with HSV-2, it does ameliorate its symptoms and decrease its severity. This would be important in vaccine trials. A small number of studies, however, found a reduced prevalence of HSV-2 antibody in patients with positive HSV-1 antibody.^9,24,26

Although western blot is the “gold standard” for the accurate detection of HSV-1 and HSV-2 antibodies, it is expensive, time consuming, and not widely available for commercial use.²⁷ In a study comparing the Gull EIA and western blotting tests, the sensitivity and specificity of the Gull EIA were found to be 95% and 96% respectively for HSV-1, and 98% and 97% respectively for HSV-2.¹⁰ In another study,¹⁹ the sensitivity and specificity values of the Gull/Meridian EIA versus western blotting were 91.9% and 98.0%, respectively. The positive and negative predictive values of the assay depend upon the prevalence of the infection among the studied population. The high prevalence of HSV antibodies among genitourinary clinic attenders may warrant its use without a confirmatory test, although equivocal results may need to be confirmed (for example, by the western blotting technique). Participants in our study were informed of the limitations of the test.

Vonau et al²⁸ assessed the knowledge and attitude of pregnant women regarding genital herpes and its serotesting, in an antenatal clinic in London. Eighty per cent of this population wanted to be tested for genital herpes antibodies. In another study performed in a GUM clinic in Leeds,²⁹ 92% wanted to know if they had been infected with genital herpes. Sixty five per cent expected testing as part of routine screening. In the last two studies, no counselling was given and patients’ understanding of the information provided was not assessed. In our study, patients’ knowledge, understanding, and opinion were assessed before and after counselling. Thus, their perception and understanding of the issues surrounding this infection and its serotesting were tested. Patients’ knowledge about the infection and its serotesting (Q1–Q8) improved significantly after counselling. Patients in our study not only expressed a strong desire to be tested but also agreed to be tested (97% of the eligible attenders). The percentage of participants supporting serotesting increased from 92% to 99% (p=0.0003), indicating increased awareness of the disease. The number of patients who were unsure of their opinion decreased from 7% to 0.9%, implying that counselling was useful in helping them to decide (table 5).

Knowledge about genital herpes is generally acknowledged to be poor among both patients and clinicians.^3,30 Another study by the authors among general practitioners in the same city³¹ demonstrated the lack of knowledge in some areas about genital herpes, especially recent information. Clinician and patient education is an essential component of management. Education of both patients and healthcare providers about genital herpes was one of the recommendations of an advisory panel, recently convened by the Centres for Disease Control and Prevention (CDC) in the United States to address genital herpes prevention.³⁰

We anticipated more post-testing problems, especially from those with positive results for HSV-2 antibody. Are we overestimating the potential psychological effect of this test? Is it patients’ or physicians’ worry? A recent study undertaken in Australia has reported a similar finding regarding the minimal psychological impact of genital herpes serotesting.³² On the contrary, 32% of our study population were reassured that there was no indication of previous herpes simplex infection—that is, negative results for HSV-1 and HSV-2 antibodies. Other possible explanations of the small number of patients who contacted the unit after discharge with post-testing concerns include “silent suffering” or difficulty in contacting the department (for example, because of moving). The long term effects of receiving the results were not assessed. Some of the patients may suffer later. Other studies are required to assess the short and long term effect of this testing in more detail.

Counselling is not only important in obtaining informed consent, but also necessary to provide patients with knowledge about the infection, combat misconception, assess their need for testing, prepare them for results, help those with positive results to come to terms with the diagnosis and, above all, to help in preventing transmission.^33–36 Counselling in our study was not controlled, as it is difficult to deny one group. However the significant improvement of knowledge, and the smooth acceptance of results without significant post-testing problems, are most likely to be due to the appropriate counselling received by the participants.

A survey carried out among delegates at the annual meeting of the International Herpes Management Forum (IHMF) in 1999 showed that 70% were in favour of judicious use of serology to test for HSV. The advisory panel, mentioned earlier, convened by the CDC in the United States,³⁰ has recommended that this testing should be routinely employed in the evaluation of genital ulceration. The panel also recommended that patients requesting STD screening or patients with symptoms suggesting an STD should be specifically informed whenever evaluation does not include this serotesting for HSV infection. Munday et al in their study³⁷ found that the type specific serology test contributed to patient management in 79% of cases with recurrent genital ulceration and was also useful for counselling patients. They proposed guidelines for using this test. Many situations emerge where knowledge of this serology could be of value.³⁸ Proper diagnosis of unrecognised cases of genital herpes (80%) is not only a fundamental right for the patient, but is also an important step in the prevention of this infection.⁶ Preventive measures are not currently perfect, but using the available tools and improving them is the way forward.