PMCC PMCC

Search tips
Search criteria

Advanced
Results 1-16 (16)
 

Clipboard (0)
None

Select a Filter Below

Journals
Year of Publication
Document Types
1.  A cognitive behavioural intervention to reduce sexually transmitted infections among gay men: randomised trial 
BMJ : British Medical Journal  2001;322(7300):1451-1456.
Objective
To determine the effectiveness of a brief cognitive behavioural intervention in reducing the incidence of sexually transmitted infections among gay men.
Design
Randomised controlled trial with 12 months' follow up.
Setting
Sexual health clinic in London.
Participants
343 gay men with an acute sexually transmitted infection or who reported having had unprotected anal intercourse in the past year.
Main outcome measures
Number of new sexually transmitted infections diagnosed during follow up and self reported incidence of unprotected anal intercourse.
Results
72% (361/499) of men invited to enter the study did so. 90% (308/343) of participants returned at least one follow up questionnaire or re-attended the clinic and requested a check up for sexually transmitted infections during follow up. At baseline, 37% (63/172) of the intervention group and 30% (50/166) of the control group reported having had unprotected anal intercourse in the past month. At 12 months, the proportions were 27% (31/114) and 32% ( 39/124) respectively (P=0.56). However, 31% (38/123) of the intervention group and 21% (35/168) of controls had had at least one new infection diagnosed at the clinic (adjusted odds ratio 1.66, 95% confidence interval 1.00 to 2.74). Considering only men who requested a check up for sexually transmitted infections, the proportion diagnosed with a new infection was 58% (53/91) for men in the intervention group and 43% (35/81) for men in the control group (adjusted odds ratio 1.84, 0.99 to 3.40). Using a regional database that includes information from 23 sexual health clinics in London, we determined that few participants had attended other sexual health clinics.
Conclusions
This behavioural intervention was acceptable and feasible to deliver, but it did not reduce the risk of acquiring a new sexually transmitted infection among these gay men at high risk. Even carefully designed interventions should not be assumed to bring benefit. It is important to evaluate their effects in randomised trials with objective clinical end points.
What is already known on this topicThe need for effective HIV prevention strategies based on reducing sexual risk behaviour remains importantFew interventions to reduce sexual risk behaviour have been rigorously evaluated using randomised controlled trialsWhat this study addsThis is the first randomised controlled trial of an intervention addressing sexual behaviour in homosexual men that uses sexually transmitted infections and self reported behaviour as end pointsThe intervention was brief and feasible to use in a busy clinic, but it did not reduce the risk of participants acquiring new infectionsThe potential for behavioural interventions to do more harm than good needs to be taken seriously
PMCID: PMC32307  PMID: 11408300
2.  Nonprobability Web Surveys to Measure Sexual Behaviors and Attitudes in the General Population: A Comparison With a Probability Sample Interview Survey 
Background
Nonprobability Web surveys using volunteer panels can provide a relatively cheap and quick alternative to traditional health and epidemiological surveys. However, concerns have been raised about their representativeness.
Objective
The aim was to compare results from different Web panels with a population-based probability sample survey (n=8969 aged 18-44 years) that used computer-assisted self-interview (CASI) for sensitive behaviors, the third British National Survey of Sexual Attitudes and Lifestyles (Natsal-3).
Methods
Natsal-3 questions were included on 4 nonprobability Web panel surveys (n=2000 to 2099), 2 using basic quotas based on age and sex, and 2 using modified quotas based on additional variables related to key estimates. Results for sociodemographic characteristics were compared with external benchmarks and for sexual behaviors and opinions with Natsal-3. Odds ratios (ORs) were used to express differences between the benchmark data and each survey for each variable of interest. A summary measure of survey performance was the average absolute OR across variables. Another summary measure was the number of key estimates for which the survey differed significantly (at the 5% level) from the benchmarks.
Results
For sociodemographic variables, the Web surveys were less representative of the general population than Natsal-3. For example, for men, the average absolute OR for Natsal-3 was 1.14, whereas for the Web surveys the average absolute ORs ranged from 1.86 to 2.30. For all Web surveys, approximately two-thirds of the key estimates of sexual behaviors were different from Natsal-3 and the average absolute ORs ranged from 1.32 to 1.98. Differences were appreciable even for questions asked by CASI in Natsal-3. No single Web survey performed consistently better than any other did. Modified quotas slightly improved results for men, but not for women.
Conclusions
Consistent with studies from other countries on less sensitive topics, volunteer Web panels provided appreciably biased estimates. The differences seen with Natsal-3 CASI questions, where mode effects may be similar, suggest a selection bias in the Web surveys. The use of more complex quotas may lead to some improvement, but many estimates are still likely to differ. Volunteer Web panels are not recommended if accurate prevalence estimates for the general population are a key objective.
doi:10.2196/jmir.3382
PMCID: PMC4275497  PMID: 25488851
Internet survey; Web survey; survey methods; sampling bias; selection bias; sexual behavior
3.  A Declining CD4 Count and Diagnosis of HIV-Associated Hodgkin Lymphoma: Do Prior Clinical Symptoms and Laboratory Abnormalities Aid Diagnosis? 
PLoS ONE  2014;9(2):e87442.
Background
The incidence of Hodgkin lymphoma (HL) among HIV-infected individuals remains unchanged since the introduction of combination antiretroviral therapy (cART). Recent epidemiological data suggest that CD4 count decline over a year is associated with subsequent diagnosis of HL. In an era of economic austerity monitoring the efficacy of cART by CD4 counts may no longer be required where CD4 count>350 cells/µl and viral load is suppressed (<50 copies/ml).
Methods
We sought to establish among our HIV outpatient cohort whether a CD4 count decline prior to diagnosis of HL, whether any decline was greater than in patients without the diagnosis, and also whether other clinical or biochemical indices were reliably associated with the diagnosis.
Results
Twenty-nine patients with a diagnosis of HL were identified. Among 15 individuals on cART with viral load <50 copies/ml the change in CD4 over 12 months preceding diagnosis of HL was −82 cells/µl (95% CI −163 to −3; p = 0.04). Among 18 matched controls the mean change was +5 cells/µl, 95% CI −70 to 80, p = 0.89). The decline in CD4 over the previous 6–12 months was somewhat greater in cases than controls (mean difference in change −55 cells/µl, 95% CI −151 to 39; p = 0.25). In 26 (90%) patients B symptoms had been present for a median of three months (range one–12) before diagnosis of HL.
Conclusions
The CD4 count decline in the 12 months prior to diagnosis of Hodgkin lymphoma among HIV-infected individuals with VL<50 copies/ml on cART was not significantly different from that seen in other fully virologically suppressed individuals in receipt of cART and who did not develop HL. All those who developed HL had B symptoms and/or new palpable lymphadenopathy, suggesting that CD4 count monitoring if performed less frequently, or not at all, among those virologically suppressed individuals with CD4 counts >350 may not have delayed diagnosis.
doi:10.1371/journal.pone.0087442
PMCID: PMC3913599  PMID: 24504076
4.  Sexual function in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) 
Lancet  2013;382(9907):1817-1829.
Summary
Background
Despite its importance to sexual health and wellbeing, sexual function is given little attention in sexual health policy. Population-based studies are needed to understand sexual function across the life course.
Methods
We undertook a probability sample survey (the third National Survey of Sexual Attitudes and Lifestyles [Natsal-3]) of 15 162 individuals aged 16–74 years who lived in Britain (England, Scotland, and Wales). Interviews were done between Sept 6, 2010, and Aug 31, 2012. We assessed the distribution of sexual function by use of a novel validated measure (the Natsal-SF), which assessed problems with individual sexual response, sexual function in a relationship context, and self-appraisal of sex life (17 items; 16 items per gender). We assess factors associated with low sexual function (defined as the lowest quintile of distribution of Natsal-SF scores) and the distribution of components of the measure. Participants reporting one or more sexual partner in the past year were given a score on the Natsal-SF (11 690 participants). 4122 of these participants were not in a relationship for all of the past year and we employed the full information maximum likelihood method to handle missing data on four relationship items.
Findings
We obtained data for 4913 men and 6777 women for the Natsal-SF. For men and women, low sexual function was associated with increased age, and, after age-adjustment, with depression (adjusted odds ratio 3·70 [95% CI 2·90–4·72] for men and 4·11 [3·36–5·04] for women) and self-reported poor health status (2·63 [1·73–3·98] and 2·41 [1·72–3·39]). Low sexual function was also associated with experiencing the end of a relationship (1·52 [1·18–1·95] and 1·77 [1·44–2·17]), inability to talk easily about sex with a partner (2·36 [1·94–2·88] and 2·82 [2·28–3·48]), and not being happy in the relationship (2·89 [2·32–3·61] and 4·10 [3·39–4·97]). Associations were also noted with engaging in fewer than four sex acts in the past 4 weeks (3·13 [2·58–3·79] and 3·38 [2·80–4·09]), having had same sex partners (2·28 [1·56–3·35] and 1·60 [1·16–2·20]), paying for sex (in men only; 2·62 [1·46–4·71]), and higher numbers of lifetime sexual partners (in women only; 2·12 [1·68–2·67] for ten or more partners). Low sexual function was also associated with negative sexual health outcomes such as experience of non-volitional sex (1·98 [1·14–3·43] and 2·18 [1·79–2·66]) and STI diagnosis (1·50 [1·06–2·11] and 1·83 [1·35–2·47]). Among individuals reporting sex in the past year, problems with sexual response were common (41·6% of men and 51·2% of women reported one or more problem) but self-reported distress about sex lives was much less common (9·9% and 10·9%). For individuals in a sexual relationship for the past year, 23·4% of men and 27·4% of women reported an imbalance in level of interest in sex between partners, and 18·0% of men and 17·1% of women said that their partner had had sexual difficulties. Most participants who did not have sex in the past year were not dissatisfied, distressed, or avoiding sex because of sexual difficulties.
Interpretation
Wide variability exists in the distribution of sexual function scores. Low sexual function is associated with negative sexual health outcomes, supporting calls for a greater emphasis on sexual function in sexual health policy and interventions.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and the Department of Health.
doi:10.1016/S0140-6736(13)62366-1
PMCID: PMC3898902  PMID: 24286787
5.  The prevalence of unplanned pregnancy and associated factors in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) 
Lancet  2013;382(9907):1807-1816.
Summary
Background
Unplanned pregnancy is a key public health indicator. We describe the prevalence of unplanned pregnancy, and associated factors, in a general population sample in Britain (England, Scotland, and Wales).
Method
We did a probability sample survey, the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3), of 15 162 men and women aged 16–74 years in Britain, including 5686 women of child-bearing age (16–44 years) who were included in the pregnancy analysis, between Sept 6, 2010, and Aug 31, 2012. We describe the planning status of pregnancies with known outcomes in the past year, and report the annual population prevalence of unplanned pregnancy, using a validated, multicriteria, multi-outcome measure (the London Measure of Unplanned Pregnancy). We set the findings in the context of secular trends in reproductive health-related events, and patterns across the life course.
Findings
9·7% of women aged 16–44 years had pregnancies with known outcome in the year before interview, of which 16·2% (95% CI 13·1–19·9) scored as unplanned, 29·0% (25·2–33·2) as ambivalent, and 54·8% (50·3–59·2) as planned, giving an annual prevalence estimate for unplanned pregnancy of 1·5% (1·2–1·9). Pregnancies in women aged 16–19 years were most commonly unplanned (45·2% [30·8–60·5]). However, most unplanned pregnancies were in women aged 20–34 years (62·4% [50·2–73·2]). Factors strongly associated with unplanned pregnancy were first sexual intercourse before 16 years of age (age-adjusted odds ratio 2·85 [95% CI 1·77–4·57], current smoking (2·47 [1·46–4·18]), recent use of drugs other than cannabis (3·41 [1·64–7·11]), and lower educational attainment. Unplanned pregnancy was also associated with lack of sexual competence at first sexual intercourse (1·90 [1·14–3·08]), reporting higher frequency of sex (2·11 [1·25–3·57] for five or more times in the past 4 weeks), receiving sex education mainly from a non-school-based source (1·84 [1·12–3·00]), and current depression (1·96 [1·10–3·47]).
Interpretation
The increasing intervals between first sexual intercourse, cohabitation, and childbearing means that, on average, women in Britain spend about 30 years of their life needing to avert an unplanned pregnancy. Our data offer scope for primary prevention aimed at reducing the rate of unplanned conceptions, and secondary prevention aimed at modification of health behaviours and health disorders in unplanned pregnancy that might be harmful for mother and child.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and the Department of Health.
doi:10.1016/S0140-6736(13)62071-1
PMCID: PMC3898922  PMID: 24286786
6.  Lifetime prevalence, associated factors, and circumstances of non-volitional sex in women and men in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) 
Lancet  2013;382(9907):1845-1855.
Summary
Background
Sexual violence is increasingly recognised as a public health issue. Information about prevalence, associated factors, and consequences for health in the population of Britain (England, Scotland, and Wales) is scarce. The third National Survey of Sexual Health Attitudes and Lifestyles (Natsal-3) is the first of the Natsal surveys to include questions about sexual violence and the first population-based survey in Britain to explore the issue outside the context of crime.
Methods
Between Sept 6, 2010, and Aug 31, 2012, we did a probability sample survey of women and men aged 16–74 years living in Britain. We asked participants about their experience of sex against their will since age 13 years and the circumstances surrounding the most recent occurrence. We explored associations between ever experiencing non-volitional sex and a range of sociodemographic, health, and behavioural factors. We used logistic regression to estimate age-adjusted odds ratios to analyse factors associated with the occurrence of completed non-volitional sex in women and men.
Findings
We interviewed 15 162 people. Completed non-volitional sex was reported by 9·8% (95% CI 9·0–10·5) of women and 1·4% (1·1–1·7) of men. Median age (interdecile range) at most recent occurrence was 18 years (14–32) for women and 16 years (13–30) for men. Completed non-volitional sex varied by family structure and, in women, by age, education, and area-level deprivation. It was associated with poor health, longstanding illness or disability, and treatment for mental health conditions, smoking, and use of non-prescription drugs in the past year in both sexes, and with binge drinking in women. Completed non-volitional sex was also associated with reporting of first heterosexual intercourse before 16 years of age, same-sex experience, more lifetime sexual partners, ever being diagnosed with a sexually transmitted infection, and low sexual function in both sexes, and, in women, with abortion and pregnancy outcome before 18 years of age. In most cases, the person responsible was known to the individual, although the nature of the relationship differed by age at most recent occurrence. Participants who were younger at interview were more likely to have told someone about the event and to have reported it to the police than were older participants.
Interpretation
These data provide the first population prevalence estimates of non-volitional sex in Britain. We showed it to be mainly an experience of young age and strongly associated with a range of adverse health outcomes in both women and men.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and the Department of Health.
doi:10.1016/S0140-6736(13)62300-4
PMCID: PMC3898964  PMID: 24286789
7.  Associations between health and sexual lifestyles in Britain: findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) 
Lancet  2013;382(9907):1830-1844.
Summary
Background
Physical and mental health could greatly affect sexual activity and fulfilment, but the nature of associations at a population level is poorly understood. We used data from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) to explore associations between health and sexual lifestyles in Britain (England, Scotland, and Wales).
Methods
Men and women aged 16–74 years who were resident in households in Britain were interviewed between Sept 6, 2010, and Aug 31, 2012. Participants completed the survey in their own homes through computer-assisted face-to-face interviews and self-interview. We analysed data for self-reported health status, chronic conditions, and sexual lifestyles, weighted to account for unequal selection probabilities and non-response to correct for differences in sex, age group, and region according to 2011 Census figures.
Findings
Interviews were done with 15 162 participants (6293 men, 8869 women). The proportion reporting recent sexual activity (one or more occasion of vaginal, oral, or anal sex with a partner of the opposite sex, or oral or anal sex or genital contact with a partner of the same sex in the past 4 weeks) decreased with age after the age of 45 years in men and after the age of 35 years in women, while the proportion in poorer health categories increased with age. Recent sexual activity was less common in participants reporting bad or very bad health than in those reporting very good health (men: 35·7% [95% CI 28·6–43·5] vs 74·8% [72·7–76·7]; women: 34·0% [28·6–39·9] vs 67·4% [65·4–69·3]), and this association remained after adjusting for age and relationship status (men: adjusted odds ratio [AOR] 0·29 [95% CI 0·19–0·44]; women: 0·43 [0·31–0·61]). Sexual satisfaction generally decreased with age, and was significantly lower in those reporting bad or very bad health than in those reporting very good health (men: 45·4% [38·4–52·7] vs 69·5% [67·3–71·6], AOR 0·51 [0·36–0·72]; women: 48·6% [42·9–54·3] vs 65·6% [63·6–67·4], AOR 0·69 [0·53–0·91]). In both sexes, reduced sexual activity and reduced satisfaction were associated with limiting disability and depressive symptoms, and reduced sexual activity was associated with chronic airways disease and difficulty walking up the stairs because of a health problem. 16·6% (95% CI 15·4–17·7) of men and 17·2% (16·3–18·2) of women reported that their health had affected their sex life in the past year, increasing to about 60% in those reporting bad or very bad health. 23·5% (20·3–26·9) of men and 18·4% (16·0–20·9) of women who reported that their health affected their sex life reported that they had sought clinical help (>80% from general practitioners; <10% from specialist services).
Interpretation
Poor health is independently associated with decreased sexual activity and satisfaction at all ages in Britain. Many people in poor health report an effect on their sex life, but few seek clinical help. Sexual lifestyle advice should be a component of holistic health care for patients with chronic ill health.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and Department of Health.
doi:10.1016/S0140-6736(13)62222-9
PMCID: PMC3898988  PMID: 24286788
8.  Changes in sexual attitudes and lifestyles in Britain through the life course and over time: findings from the National Surveys of Sexual Attitudes and Lifestyles (Natsal) 
Lancet  2013;382(9907):1781-1794.
Summary
Background
Sexual behaviour and relationships are key components of wellbeing and are affected by social norms, attitudes, and health. We present data on sexual behaviours and attitudes in Britain (England, Scotland, and Wales) from the three National Surveys of Sexual Attitudes and Lifestyles (Natsal).
Methods
We used a multistage, clustered, and stratified probability sample design. Within each of the 1727 sampled postcode sectors for Natsal-3, 30 or 36 addresses were randomly selected and then assigned to interviewers. To oversample individuals aged 16–34 years, we randomly allocated addresses to either the core sample (in which individuals aged 16–74 years were eligible) or the boost sample (in which only individuals aged 16–34 years were eligible). Interviewers visited all sampled addresses between Sept 6, 2010, and Aug 31, 2012, and randomly selected one eligible individual from each household to be invited to participate. Participants completed the survey in their own homes through computer-assisted face-to-face interviews and self-interview. We analysed data from this survey, weighted to account for unequal selection probabilities and non-response to correct for differences in sex, age group, and region according to 2011 Census figures. We then compared data from participants aged 16–44 years from Natsal-1 (1990–91), Natsal-2 (1999–2001), and Natsal-3.
Findings
Interviews were completed with 15 162 participants (6293 men, 8869 women) from 26 274 eligible addresses (57·7%). 82·1% (95% CI 81·0–83·1%) of men and 77·7% (76·7–78·7%) of women reported at least one sexual partner of the opposite sex in the past year. The proportion generally decreased with age, as did the range of sexual practices with partners of the opposite sex, especially in women. The increased sexual activity and diversity reported in Natsal-2 in individuals aged 16–44 years when compared with Natsal-1 has generally been sustained in Natsal-3, but in men has generally not risen further. However, in women, the number of male sexual partners over the lifetime (age-adjusted odds ratio 1·18, 95% CI 1·08–1·28), proportion reporting ever having had a sexual experience with genital contact with another woman (1·69, 1·43–2·00), and proportion reporting at least one female sexual partner in the past 5 years (2·00, 1·59–2·51) increased in Natsal-3 compared with Natsal-2. While reported number of occasions of heterosexual intercourse in the past 4 weeks had reduced since Natsal-2, we recorded an expansion of heterosexual repertoires—particularly in oral and anal sex—over time. Acceptance of same-sex partnerships and intolerance of non-exclusivity in marriage increased in men and women in Natsal-3.
Interpretation
Sexual lifestyles in Britain have changed substantially in the past 60 years, with changes in behaviour seeming greater in women than men. The continuation of sexual activity into later life—albeit reduced in range and frequency—emphasises that attention to sexual health and wellbeing is needed throughout the life course.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and the Department of Health.
doi:10.1016/S0140-6736(13)62035-8
PMCID: PMC3899021  PMID: 24286784
9.  Prevalence, risk factors, and uptake of interventions for sexually transmitted infections in Britain: findings from the National Surveys of Sexual Attitudes and Lifestyles (Natsal) 
Lancet  2013;382(9907):1795-1806.
Summary
Background
Population-based estimates of prevalence, risk distribution, and intervention uptake inform delivery of control programmes for sexually transmitted infections (STIs). We undertook the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) after implementation of national sexual health strategies, and describe the epidemiology of four STIs in Britain (England, Scotland, and Wales) and the uptake of interventions.
Methods
Between Sept 6, 2010 and Aug 31, 2012, we did a probability sample survey of 15 162 women and men aged 16–74 years in Britain. Participants were interviewed with computer-assisted face-to-face and self-completion questionnaires. Urine from a sample of participants aged 16–44 years who reported at least one sexual partner over the lifetime was tested for the presence of Chlamydia trachomatis, type-specific human papillomavirus (HPV), Neisseria gonorrhoeae, and HIV antibody. We describe age-specific and sex-specific prevalences of infection and intervention uptake, in relation to demographic and behavioural factors, and explore changes since Natsal-1 (1990–91) and Natsal-2 (1999–2001).
Findings
Of 8047 eligible participants invited to provide a urine sample, 4828 (60%) agreed. We excluded 278 samples, leaving 4550 (94%) participants with STI test results. Chlamydia prevalence was 1·5% (95% CI 1·1–2·0) in women and 1·1% (0·7–1·6) in men. Prevalences in individuals aged 16–24 years were 3·1% (2·2–4·3) in women and 2·3% (1·5–3·4) in men. Area-level deprivation and higher numbers of partners, especially without use of condoms, were risk factors. However, 60·4% (45·5–73·7) of chlamydia in women and 43·3% (25·9–62·5) in men was in individuals who had had one partner in the past year. Among sexually active 16–24-year-olds, 54·2% (51·4–56·9) of women and 34·6% (31·8–37·4) of men reported testing for chlamydia in the past year, with testing higher in those with more partners. High-risk HPV was detected in 15·9% (14·4–17·5) of women, similar to in Natsal-2. Coverage of HPV catch-up vaccination was 61·5% (58·2–64·7). Prevalence of HPV types 16 and 18 in women aged 18–20 years was lower in Natsal-3 than Natsal-2 (5·8% [3·9–8·6] vs 11·3% [6·8–18·2]; age-adjusted odds ratio 0·44 [0·21–0·94]). Gonorrhoea (<0·1% prevalence in women and men) and HIV (0·1% prevalence in women and 0·2% in men) were uncommon and restricted to participants with recognised high-risk factors. Since Natsal-2, substantial increases were noted in attendance at sexual health clinics (from 6·7% to 21·4% in women and from 7·7% to 19·6% in men) and HIV testing (from 8·7% to 27·6% in women and from 9·2% to 16·9% in men) in the past 5 years.
Interpretation
STIs were distributed heterogeneously, requiring general and infection-specific interventions. Increases in testing and attendance at sexual health clinics, especially in people at highest risk, are encouraging. However, STIs persist both in individuals accessing and those not accessing services. Our findings provide empirical evidence to inform future sexual health interventions and services.
Funding
Grants from the UK Medical Research Council and the Wellcome Trust, with support from the Economic and Social Research Council and the Department of Health.
doi:10.1016/S0140-6736(13)61947-9
PMCID: PMC3899025  PMID: 24286785
10.  Evaluation of Respondent-Driven Sampling 
Epidemiology (Cambridge, Mass.)  2012;23(1):138-147.
Background
Respondent-driven sampling is a novel variant of link-tracing sampling for estimating the characteristics of hard-to-reach groups, such as HIV prevalence in sex-workers. Despite its use by leading health organizations, the performance of this method in realistic situations is still largely unknown. We evaluated respondent-driven sampling by comparing estimates from a respondent-driven sampling survey with total-population data.
Methods
Total-population data on age, tribe, religion, socioeconomic status, sexual activity and HIV status were available on a population of 2402 male household-heads from an open cohort in rural Uganda. A respondent-driven sampling (RDS) survey was carried out in this population, employing current methods of sampling (RDS sample) and statistical inference (RDS estimates). Analyses were carried out for the full RDS sample and then repeated for the first 250 recruits (small sample).
Results
We recruited 927 household-heads. Full and small RDS samples were largely representative of the total population, but both samples under-represented men who were younger, of higher socioeconomic status, and with unknown sexual activity and HIV status. Respondent-driven-sampling statistical-inference methods failed to reduce these biases. Only 31%-37% (depending on method and sample size) of RDS estimates were closer to the true population proportions than the RDS sample proportions. Only 50%-74% of respondent-driven-sampling bootstrap 95% confidence intervals included the population proportion.
Conclusions
Respondent-driven sampling produced a generally representative sample of this well-connected non-hidden population. However, current respondent-driven-sampling inference methods failed to reduce bias when it occurred. Whether the data required to remove bias and measure precision can be collected in a respondent-driven sampling survey is unresolved. Respondent-driven sampling should be regarded as a (potentially superior) form of convenience-sampling method, and caution is required when interpreting findings based on the sampling method.
doi:10.1097/EDE.0b013e31823ac17c
PMCID: PMC3277908  PMID: 22157309
11.  Combining Multiple Imputation and Inverse-Probability Weighting 
Biometrics  2012;68(1):129-137.
Summary
Two approaches commonly used to deal with missing data are multiple imputation (MI) and inverse-probability weighting (IPW). IPW is also used to adjust for unequal sampling fractions. MI is generally more efficient than IPW but more complex. Whereas IPW requires only a model for the probability that an individual has complete data (a univariate outcome), MI needs a model for the joint distribution of the missing data (a multivariate outcome) given the observed data. Inadequacies in either model may lead to important bias if large amounts of data are missing. A third approach combines MI and IPW to give a doubly robust estimator. A fourth approach (IPW/MI) combines MI and IPW but, unlike doubly robust methods, imputes only isolated missing values and uses weights to account for remaining larger blocks of unimputed missing data, such as would arise, e.g., in a cohort study subject to sample attrition, and/or unequal sampling fractions. In this article, we examine the performance, in terms of bias and efficiency, of IPW/MI relative to MI and IPW alone and investigate whether the Rubin’s rules variance estimator is valid for IPW/MI. We prove that the Rubin’s rules variance estimator is valid for IPW/MI for linear regression with an imputed outcome, we present simulations supporting the use of this variance estimator in more general settings, and we demonstrate that IPW/MI can have advantages over alternatives. IPW/MI is applied to data from the National Child Development Study.
doi:10.1111/j.1541-0420.2011.01666.x
PMCID: PMC3412287  PMID: 22050039
Marginal model; Missing at random; Survey weighting; 1958 British Birth Cohort
12.  Quantifying, displaying and accounting for heterogeneity in the meta-analysis of RCTs using standard and generalised Q statistics 
Background
Clinical researchers have often preferred to use a fixed effects model for the primary interpretation of a meta-analysis. Heterogeneity is usually assessed via the well known Q and I2 statistics, along with the random effects estimate they imply. In recent years, alternative methods for quantifying heterogeneity have been proposed, that are based on a 'generalised' Q statistic.
Methods
We review 18 IPD meta-analyses of RCTs into treatments for cancer, in order to quantify the amount of heterogeneity present and also to discuss practical methods for explaining heterogeneity.
Results
Differing results were obtained when the standard Q and I2 statistics were used to test for the presence of heterogeneity. The two meta-analyses with the largest amount of heterogeneity were investigated further, and on inspection the straightforward application of a random effects model was not deemed appropriate. Compared to the standard Q statistic, the generalised Q statistic provided a more accurate platform for estimating the amount of heterogeneity in the 18 meta-analyses.
Conclusions
Explaining heterogeneity via the pre-specification of trial subgroups, graphical diagnostic tools and sensitivity analyses produced a more desirable outcome than an automatic application of the random effects model. Generalised Q statistic methods for quantifying and adjusting for heterogeneity should be incorporated as standard into statistical software. Software is provided to help achieve this aim.
doi:10.1186/1471-2288-11-41
PMCID: PMC3102034  PMID: 21473747
13.  Early Predictors of Mortality from Pneumocystis jirovecii Pneumonia in HIV-Infected Patients: 1985–2006 
Background
Pneumocystis jirovecii pneumonia (PCP) remains the leading cause of opportunistic infection among human immunodeficiency virus (HIV)–infected persons. Previous studies of PCP that identified case-fatality risk factors involved small numbers of patients, were performed over few years, and often focused on patients who were admitted to the intensive care unit.
Objective
The objective of this study was to identify case-fatality risk factors present at or soon after hospitalization among adult HIV-infected patients admitted to University College London Hospitals (London, United Kingdom) from June 1985 through June 2006.
Patients and Methods
We performed a review of case notes for 494 consecutive patients with 547 episodes of laboratory-confirmed PCP.
Results
Overall mortality was 13.5%. Mortality was 10.1% for the period from 1985 through 1989, 16.9% for the period from 1990 through June 1996, and 9.7% for the period from July 1996 through 2006 (P = .142). Multivariate analysis identified factors associated with risk of death, including increasing patient age (adjusted odds ratio [AOR], 1.54; 95% confidence interval [CI], 1.11–2.23; P = .011), subsequent episode of PCP (AOR, 2.27; 95% CI, 1.14–4.52; P = .019), low hemoglobin level at hospital admission (AOR, 0.70; 95% CI, 0.60–0.83; P < .001), low partial pressure of oxygen breathing room air at hospital admission (AOR, 0.70; 95% CI, 0.60–0.81; P < .001), presence of medical comorbidity (AOR, 3.93; 95% CI, 1.77–8.72; P = .001), and pulmonary Kaposi sarcoma (AOR, 6.95; 95% CI, 2.26–21.37; P =.001). Patients with a first episode of PCP were sicker (mean partial pressure of oxygen at admission ± standard deviation, 9.3 ± 2.0 kPa) than those with a second or third episode of PCP (mean partial pressure of oxygen at admission ± standard deviation, 9.9 ± 1.9 kPa; P =.008), but mortality among patients with a first episode of PCP (12.5%) was lower than mortality among patients with subsequent episodes of PCP (22.5%) (P = .019). No patient was receiving highly active antiretroviral therapy before presentation with PCP, and none began highly active antiretroviral therapy during treatment of PCP.
Conclusions
Mortality risk factors for PCP were identifiable at or soon after hospitalization. The trend towards improved outcome after June 1996 occurred in the absence of highly active antiretroviral therapy.
doi:10.1086/526778
PMCID: PMC2735405  PMID: 18190281
14.  Consistency in reporting sensitive sexual behaviours in Britain: change in reporting bias in the second and third National Surveys of Sexual Attitudes and Lifestyles (Natsal-2 and Natsal-3) 
Objectives
Britain's second National Survey of Sexual Attitudes and Lifestyles (Natsal-2) was conducted in 1999–2001 and the third (Natsal-3) was conducted in 2010–2012 to update prevalence estimates of sexual behaviours and assess changes over time. We investigated whether there was a change in reporting bias between these two cross-sectional surveys.
Methods
We analysed data from the ‘common birth cohort’ of participants born during 1956–1983, who were eligible to take part in Natsal-2 (n=10 764) and Natsal-3 (n=6907). We compared estimates for outcomes that occurred before Natsal-2 and expected these to be consistent between surveys if no change in reporting bias had occurred.
Results
A greater proportion of non-white men and women were in Natsal-3 consistent with demographic changes in Britain. Reporting behaviours was largely consistent between surveys for men. Fewer women in Natsal-3 reported early first intercourse or having child(ren) before age 20; they were also more likely to report not discussing sex with their parents at age 14. Men and women in Natsal-3 were more likely to report tolerant attitudes towards same-sex partnerships but less tolerance towards unfaithfulness in marriage and one-night-stands.
Conclusions
We found little evidence of change in reporting bias among men since Natsal-2. Among women, a modest change in reporting bias was observed for a small number of experiences, possibly due to changes in participation, social acceptability and methodological differences between surveys. Changes in the reporting of sexual behaviours and attitudes over time observed in the wider Natsal-3 study are therefore likely to largely reflect real changes in the population.
doi:10.1136/sextrans-2013-051360
PMCID: PMC3932741  PMID: 24277882
Epidemiology (General); Sexual Behaviour; Statistics
15.  Methodology of the third British National Survey of Sexual Attitudes and Lifestyles (Natsal-3) 
Background
Data from the first two National Surveys of Sexual Attitudes and Lifestyles, carried out in 1990–1991 (Natsal-1) and 1999–2001 (Natsal-2), have been extensively used to inform sexual health policy in Britain over the past two decades. Natsal-3 was carried out from September 2010 to August 2012 in order to provide up-to-date measures of sexual lifestyles and to extend the scope of the previous studies by including an older age group (up to 74 years), an extended range of topics and biological measures.
Methods
We describe the methods used in Natsal-3, which surveyed the general population in Britain aged 16–74 years (with oversampling of younger adults aged 16–34 years).
Results
Overall, 15 162 interviews were completed, with a response rate of 57.7% and a cooperation rate of 65.8%. The response rate for the boost sample of ages 16–34 years was 64.8%, only marginally lower than the 65.4% achieved for Natsal-2, which surveyed a similar age range (16–44). The data were weighted by age, gender and region to reduce possible bias. Comparisons with census data show the weighted sample to provide good representation on a range of respondent characteristics. The interview involved a combination of face-to-face and self-completion components, both carried out on computer. Urine samples from 4550 sexually-experienced participants aged 16–44 years were tested for a range of STIs. Saliva samples from 4128 participants aged 18–74 years were tested for testosterone.
Conclusions
Natsal-3 provides a high quality dataset that can be used to examine trends in sexual attitudes and behaviours over the past 20 years.
doi:10.1136/sextrans-2013-051359
PMCID: PMC3933071  PMID: 24277881
SEXUAL BEHAVIOUR; SEXUAL HEALTH; SOCIAL SCIENCE; STATISTICS
16.  Methods for Observed-Cluster Inference When Cluster Size Is Informative: A Review and Clarifications 
Biometrics  2014;70(2):449-456.
Clustered data commonly arise in epidemiology. We assume each cluster member has an outcome Y and covariates . When there are missing data in Y, the distribution of Y given in all cluster members (“complete clusters”) may be different from the distribution just in members with observed Y (“observed clusters”). Often the former is of interest, but when data are missing because in a fundamental sense Y does not exist (e.g., quality of life for a person who has died), the latter may be more meaningful (quality of life conditional on being alive). Weighted and doubly weighted generalized estimating equations and shared random-effects models have been proposed for observed-cluster inference when cluster size is informative, that is, the distribution of Y given in observed clusters depends on observed cluster size. We show these methods can be seen as actually giving inference for complete clusters and may not also give observed-cluster inference. This is true even if observed clusters are complete in themselves rather than being the observed part of larger complete clusters: here methods may describe imaginary complete clusters rather than the observed clusters. We show under which conditions shared random-effects models proposed for observed-cluster inference do actually describe members with observed Y. A psoriatic arthritis dataset is used to illustrate the danger of misinterpreting estimates from shared random-effects models.
doi:10.1111/biom.12151
PMCID: PMC4312901  PMID: 24479899
Bridge distribution; Immortal cohort inference; Informative missingness; Missing not at random; Mortal cohort inference; Semi-continuous data

Results 1-16 (16)