Background Open access to sexual health services may be inefficient if there are substantial unpredictable fluctuations in presentations. Our aim was to determine whether the number of presentations over the last 11 years was associated with certain factors.
Methods This study involved all individuals presenting to Melbourne Sexual Health Centre (MSHC) from 2002 to 2012. The outcome measure was the number of presentations during a clinical session (half day).
Results There were 270 070 presentations to the clinic among 86 717 individuals. The factors associated with the largest difference in mean presentations per session were morning or afternoon (60 vs 51 per session), days of the week (57–67 per session), months of the year (93–112 per day), year (77–131 per day), maximum temperatures of <15°C vs ≥30°C (56–62 per morning session) and 5 working days after holiday periods (61 vs 54). A multiple linear regression model using these factors explained 64% of the variation in attendances per session. Peak attendance rates (>90th centile) were also strongly correlated with these same variables. Higher-risk heterosexuals (≤25 years of age) attended more commonly in the afternoons (37% of heterosexuals) than in the mornings (30%). No factor other than year of attendance substantially influenced the proportion of higher-risk men who have sex with men (MSM) (≥10 partners per year) who attended.
Conclusions A considerable proportion of the variability in presentations was explained by known factors that could predict client presentations to sexual health services and therefore allow optimal allocation of resources to match demand.
- Epidemiology (General)
- Service Delivery
- Sexual Health
Statistics from Altmetric.com
Access to health services for testing and treatment of bacterial sexually transmitted infections (STI) shortens their duration in addition to providing access to other services such as counselling and partner notification.1 ,2 The higher the risk of those accessing these services the greater the effect will be on the community prevalence of STI.2 However, providing effective STI services is hampered by increasingly limited resources and funding for public STI clinics in many countries, including the closure of STI clinics in the USA.3 These pressures have resulted in many services developing strategies to improve clinical efficiency and better manage their clinical load.4 Many sexual health clinics have moved away from booked appointments towards open access models that allow triage of high-risk clients on the day of presentation.5 This strategy, however, may be inefficient if there are substantial fluctuations in presentations that leave the service with insufficient clients on some days and too many on other days.
One approach to help manage the problem of large variations in the number of daily client presentations would be to predict attendances in advance. If there were factors that predicted attendances, then this would allow the staffing capacity of clinics to be varied accordingly. We were unable to find any published studies addressing this issue in sexual health services. Calendar and weather variables have predicted patient attendances to an emergency department.6–8 In one USA study, summer and weather extremes were associated with 5–20% fewer paediatric emergency department attendances.6 In another study from USA, presentations to hospital emergency department were higher on a Monday and lower on days with rain.7 The authors could accurately predict patient flow on a weekly or seasonal basis, thus enabling more informed staffing decision to be made.7 Another study assessed similar variables and found that they explained 84% of daily variance in presentations to walk-in clinics in an emergency department.8
The Melbourne Sexual Health Centre (MSHC) operates a walk-in triage clinic that experienced considerable variation in daily attendances. Our aim was to determine whether attendances over the last 10 years were associated with a number of variables.
Materials and methods
This study included all individuals presenting to MSHC from 5 February 2002 to 31 December 2012. MSHC is the main public sexual health clinic in Victoria, Australia. The clinic provides about 35 000 consultations annually for free HIV and STI testing. The clinic operates a walk-in service where all new clients are triaged in by a registered nurse. Clients triaged in were seen in the clinic on same day or triaged out to another health service. Clients may be triaged out to other health services (primarily to general practitioners) because the clinic is unable to see them on that day or because they are considered at low risk of STIs. Patients who attend in the morning are not triaged out to return on the same day. The morning clinics run from 9:00 to 12:30, and the afternoon clinic from 13.30 to 17:00. We defined a morning presentation as arriving to the clinic before 13:00 (time stamp on electronic registration) and afternoon presentations are those arriving after 13:00. MSHC has four morning and five afternoon clinics (no Friday morning clinic). A limited number of clients are given appointments for up to 4 weeks in the future for review of a current clinical condition. Appointments are not included in this analysis.
The software at MSHC records the age and sex of every individual presenting to the centre. For those who are triaged into the service, basic information on sexual risks such as number of male and female sexual partners in the last 12 months are collected.
The data on daily temperature and daily rainfall were obtained from the Bureau of Meteorology.9 ,10 The data on public holidays, school vacations and university holidays were obtained from the Department of Education and Early Childhood Development of the Victorian government, Australia, and the University of Melbourne, respectively.11 ,12 Dates of major gay festivals such as ‘Midsumma’ and ‘Sydney Mardi Gras’ and music festival ‘Big Day Out’ were obtained from their respective websites.13–15 ‘Midsumma’ is a festival that celebrates the pride and diversity of Victorian's lesbian, gay and allied communities and is recognised as a principal arts and cultural festival.13 The ‘Sydney Mardi Gras’ is an annual festival held in Sydney, a city about 900 km north of Melbourne, which includes a variety of events such as Sydney Mardi Gras parade and party, Bondi Beach Drag races, harbour party, academic discussions, Mardi Gras film festival and fair day with the participation of lesbian, gay, bisexual and transgender communities from all states in Australia as well as international communities.14 The ‘Big Day Out’ is a music festival held late January each year in all five states in Australia. It has seven or eight stages accommodating popular contemporary rock music, electronic music, mainstream international acts and local acts.15
For the purposes of the analysis, we defined a presentation as a person who presented through the walk-in triage system. Individuals with booked appointments were not included. We defined men who have sex with men (MSM) as a man who has had sex with another man in the previous 12 months while a high-risk MSM is defined as a man who has had sex with 10 or more male partners in the previous 12 months. A heterosexual client was defined as a client who has had sex with opposite sex partners in the previous 12 months. A young heterosexual client is defined as a heterosexual client who was 25 years of age or less at first visit to MSHC because this is the age cut-off that many countries use for chlamydia screening. Bisexual men were coded as MSM for this analysis.
Five working days immediately after a public holiday or a university holiday were categorised as ‘five working days after a holiday’ and similarly five working days immediately after any festival out of major festivals considered were categorised as ‘five working days after each festival’. We chose 5 days because we felt that it would include those worried about a recent sexual risk and the short incubation period of urethral gonorrhoea. Most chlamydial and other sites of gonococcal infections in men are asymptomatic, and gonorrhoea in women is uncommon in Victoria.
The data were analysed using SPSS V.21. The mean number of attendances, SD and 10th and 90th centiles were calculated for different time variables, including day of week, month of year, year and proximity to a holiday or festival. T test and analysis of variance (ANOVA) test were appropriately used to investigate any association of the mean number of attendances with each different time variable. χ2 test and χ2 test for trend were used to investigate associations between MSM or heterosexuals and different variables. Multiple linear regression was used to investigate the independent effect of each time variable on the number of attendances. To investigate factors associated with marked fluctuations in number of attendances, we calculated a binary variable where the number of attendances was recoded as either above the 90th centile of number of attendances per session and for each year separately. Logistic regression was used to calculate ORs and 95% CIs for factors associated with having over the 90th centile of attendances in a session for each year. We chose a 90% cut-off to investigate whether the associations of a marked increase in attendances (ie, 90% or greater) were similar to the associations for all attendances. The meteorological variables were categorised into differences that were likely to be subjectively different. Where data were missing (eg, temperature or rainfall), the other data from this day were excluded from the analysis.
Over the study period, there were 270 070 presentations to the clinic among 86 717 individuals displayed on a scatter plot in figure 1. There were 155 290 male presentations among 50 533 individual men, 113 969 female presentations among 36 024 individual women and 811 presentations among 160 individual transsexuals. The mean and 10th and 90th centiles are shown in table 1. There was a steady rise in the number of annual presentations over the 11 years. There were substantial and statistically significant differences between presentations to the clinic on different days of the week, months of the year, years, holiday periods and weather patterns. The factors associated with the largest difference in mean presentations per session were morning or afternoon (60 vs 51 per session), days of the week (57–67 per session) and months of the year (93–112 per day), year (77–131 per day), maximum temperatures of <15°C vs ≥30°C (56–62 per morning session) and five working days after holiday periods (61 vs 54).
Online supplementary table S1 compares consultations among heterosexuals 25 years or younger and those older than 25 years. The proportion of heterosexuals 25 years or younger was substantially greater in the afternoons than in the mornings (37% vs 30%) and the month of January but varied little with other factors.
Online supplementary table S2a compares consultations in men by MSM status. The proportion of men seen for consultations in the clinic who were MSM was not substantially different by the different factors other than year. In line with clinic policy, to attract MSM to the clinic, the proportion of clinic consultations for MSM rose from 24% to 52%.
Online supplementary table S2b compares MSM with 10 or more partners in the last 12 months to those with fewer partners. The proportion of MSM with 10 or more partners in the last 12 months was not substantially different by the different factors other than year. The proportion of MSM with more than 10 partners has fallen over time as the number of MSM has increased about sixfold.
Table 2 describes the linear regression model of the independent effect of factors on the number of attendances. The factors that were significantly associated with the number of presentations to MSHC were the year of attendance, the time of day (morning or afternoon sessions), the day of the week (Monday, Tuesday and Friday), month of the year (January to May) and the 5 days after public or university holidays. This model explained 64% of the variation in presentations to MSHC.
Online supplementary table S3 shows a logistic regression model for ORs for a session having more than the 90th centile of presentations in a given year. The associations of a session with more than the 90th centile of presentations were day of the week, month of the year and up to five working days after a public holiday. Sessions with substantially more presentations per session in the 90th centile were Friday (41% ≥90th centile), January (19% ≥90th centile) and those on the first 5 days after holidays (18% ≥90th centile).
In our study, we found a number of factors were associated with substantial differences in the number of presentations to the clinic during different sessions (half days). Taken collectively, these factors explained 64% of the variation in the number of attendances for a particular session. We found that except of younger heterosexuals being seen more commonly in the afternoon, none of the factors substantially influenced the risk profile of heterosexuals or MSM seen in the clinic. This is the first study of this type undertaken in sexual health services and has important implications locally and potentially at other services that operate walk-in systems because it will allows more efficient allocation of resources. Although some of our findings will not be necessarily transferrable to different climatic conditions, other factors, such as time of the day or day of the week, may be generalisable and help guide the provision of effective clinical services in a walk-in environment. We will use these data to maximise the ability of our clinic to see and treat individuals at risk of STI as access to services is an important element of effective STI control.
Our study had a number of weaknesses that need to be considered when interpreting this data. First, the study was undertaken in a single sexual health service and it is possible that the findings of this study are not reflective of the overall client presentations to other sexual health services in other parts of the world. As this is the first study of its type, it is not possible to test this hypothesis, except that there are some consistent findings that are plausible. Specifically, demand is greater after periods of reduced services (eg, on a Monday or Friday afternoon (no clinic Friday morning)) or after leisure periods (eg, after holiday periods). Second, this was an observational study and is liable to a number of biases associated with such studies. Specifically it is possible that there are other factors that may be explaining the variations in consultations that we did not measure, although the observation that the factors we measured explained 64% of the variance in attendances suggests we identified most factors. Third, our analysis treated every attendance as independent of each other and did not include an analysis that took account of the same individual attending on multiple times. This may have overestimated the significance, but we note that p values were usually highly significant.
We specifically did not assess whether the rates of STI varied in our analysis for a number of reasons. First, on individuals who presented but were not seen or tested, we did not have information on whether or not they had STI. Second, over the period of the analysis, there were very substantial differences in the prevalence of STI among individuals presenting to the centre. For example, genital warts have fallen dramatically following the successful human papillomavirus vaccine,16 syphilis and gonorrhoea have risen substantially and finally chlamydia testing has increased greatly over this time.17 We therefore chose to use well-established surrogates of risk such as age in heterosexuals and number of partners in MSM.
We were not able to find other studies that had assessed associations of presentations to sexual health services, although a UK study found significant increases in STI diagnoses and HIV tests nationally were more common the first and third quarter of the year, one of which corresponds to the summer holiday period in Australia.18 There were a number of studies that assessed attendances to non-sexual health services. In three USA studies of emergency departments or walk-in clinics, season and weather patterns were consistently associated with attendances.6–8 In the USA, during 1975–1976, the winter was associated with greater demand for paediatric emergency department use in a total sample of 8470 presentations to their department.6 In this study, they suggested that the main reason for this was change in disease prevalence. In another study from the USA, year of attendance, month of the year and day of the week were identified as three key determinants of presentations to a large ambulatory walk-in care service of a teaching hospital. Like our study, this study also found a considerable proportion of the variation in presentations was explained by the model (77%).7 Another similar study found that 84% of daily unscheduled presentations to walk-in clinic of Veteran Affairs Medical Center could be explained by variables such as season, week of the month, day of the week, holidays and check delivery days and weather variables such as high temperature and snowfall.7 Staffing according to regression model prediction could have reduced overstaffing from 59% to 15% of days even though it would have increased understaffing from 2% to 18% of days.8
We specifically have not commented on the statistical significance of many of our findings but have rather commented on whether there were substantial differences in the variability of attendances. This is because with over a quarter of a million attendances even very small differences were statistically significant, but from the clinical perspective, relatively meaningless.
Our study has a number of strengths. It has the largest sample size of any study to date (270 070 presentations) and includes data over an 11-year period. It also includes substantial epidemiological data that allowed us to determine not only whether presentations were different but also whether the risk profile of individuals attending the centre also varied. As a result of these data, we are configuring our staff rosters to take account of the likely changes in presentations and moving review appointments to quieter times of the week. If this allows us to screen and treat a greater number of patients, then it may reduce the community prevalence of STI at no extra cost.
A considerable proportion (64%) of the variability in presentations to an Australian sexual health services can be explained by known factors.
Higher-risk heterosexuals (≤25 years of age) in Melbourne tend to attend sexual health services more commonly in the afternoon sessions.
No factor other than year of attendance influences the proportion of higher-risk men who have sex with men attending an Australian sexual health service.
The authors gratefully acknowledge A Afrizal and the staff at MSHC who assisted in many ways with this project. We are grateful to the Ministry of Health, Sri Lanka, for sponsoring the fellowship of NG.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Handling editor Jackie A Cassell
Contributors CKF conceived the study. NG and CKF wrote the manuscript and analysed the data together with JSH and ML. All authors provided intellectual inputs into the design and manuscript.
Funding NG receives a fellowship from the Ministry of Health, Sri Lanka. There was no specific funding for other aspects of the study.
Competing interests None.
Ethics approval Ethical approval for this study was granted by the Alfred Hospital Human Research Committee (number 22/13).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.