Article Text

Download PDFPDF

Original article
The utility of short message service (SMS) texts to remind patients at higher risk of STIs and HIV to reattend for testing: a controlled before and after study
  1. Jessica Burton,
  2. Gary Brook,
  3. John McSorley,
  4. Siobhan Murphy
  1. Department of Sexual Health and HIV, Central Middlesex Hospital, North West London Hospitals NHS Trust, London, UK
  1. Correspondence to Dr Gary Brook, Patrick Clements Clinic, Central Middlesex Hospital, Acton Lane, London NW10 7NS, UK; gary.brook{at}nhs.net

Abstract

Background Patients attending for sexually transmitted infection (STI)/HIV testing may be at continuing risk of infection and advised to return for retesting at a later date.

Objectives To measure the impact of short message service (SMS) text reminders on the reattendance rates of patients who require repeat STI testing.

Methods Reattendance rates were measured for two groups of higher risk patients: those listed for routine SMS text reminders in 2012 and a control group of patients from 2011 with the same risk profile who had not received any active recall. Reattendance was counted if it was within 4 months of the end of the episode of care.

Results Reattendance rates were not statistically different between the text group 32% (89/274) and the control group 35% (92/266). Reattendance also was not statistically different between the text and control groups respectively in patients with the following risks: recent chlamydia 43/121 (36%) versus 41/123 (33%), recent gonorrhoea 4/21 (19%) versus 7/21 (33%), recent emergency contraception 27/60 (45%) versus 25/56 (45%) and other risks 7/27 (26%) versus 9/26 (35%). High rates of STIs were found in patients who reattended in both the text group (13/90, 14%) and control group (15/91, 17%) and at even higher rates at reattendance if the reason for recall was chlamydia infection at the initial visit: 9/43 (21%) in the text group and 10/41 (24%) in the control group.

Conclusions SMS texts sent as reminders to patients at higher risk of STIs and HIV did not increase the reattendance rate, when compared with standard advice, in this service which already has a high reattendance rate. STI rates were high in those patients who reattended.

  • Chlamydia Infection
  • Service Delivery
  • Sexual Health
  • Contraception
  • Health Serv Research
View Full Text

Statistics from Altmetric.com

Introduction

Background knowledge

An important risk factor for testing positive for a sexually transmitted infection (STI) is having tested positive for that infection previously.1–3 Several guidelines recommend routine retesting after being diagnosed with an STI. However, the time period after which retesting should occur varies.1–3 Different methods can be used to recall patients for retesting. The simplest and least expensive method is to advise patients to retest at the time of the original diagnosis but this has not been shown to be particularly effective.4

Another recall method is to use short message service (SMS) text messaging. In Sydney, Australia, SMS text messaging to patients’ mobile ‘phones significantly increased the proportion of patients who retested after a positive chlamydia diagnosis from 21% to 30%’.5 Another Australian sexual healthcare study found that reattendance rates following a chlamydia diagnosis increased from 6.3% to 28.1% when a text reminder was sent.6 A payment incentive in the SMS test message did not make a significant difference.6

Local problem

Patients with acute STIs, sex workers, men who have sex with men (MSM) and women receiving emergency contraception are all at higher risk of future STIs.1–3 There is therefore a strong argument that, as with patients with chlamydia, these other higher risk patients should be actively recalled for repeat STI testing.

Study question

We sought to see if text recall had any utility in improving reattendance in higher risk patients.

Methods

Setting

This clinic offers both a walk-in and booked appointment service. It has approximately 15 000 sexual health attendances each year, with a high rate of STIs and HIV and a large ethnic diversity including >50% of attendees being black British/Caribbean/African.

Planning the intervention

Before September 2012, the clinic practice was to advise higher risk patients to return for repeat testing but not to send routine reminders. In September 2012, the clinic introduced SMS text reminders for patients at risk of STIs and those in the window period for HIV (ie, the period between when someone is potentially exposed to HIV and when that organism would be reliably detected in a blood test, which is 3 months). Higher risk patients included patients diagnosed with chlamydia, gonorrhoea, acute viral hepatitis or syphilis and women receiving emergency contraception, commercial sex workers, MSM and those in the window period for HIV. These patients were booked into a SMS text reminder virtual clinic if the patient consented. A text was sent after 6 weeks in the majority (80%) with a range of 2–12 weeks. This was quick and easy to do using our electronic patient records. The text message was: ‘It is time for you to have a routine test. Walk-in during opening hours or ring xxxxxxx for an appointment. Do not text back. From CMH’. This message is similar to one used successfully elsewhere.7

Planning the study of the intervention

This was a controlled before and after study.

Intervention group: patients at higher risk of STIs and who had been listed for SMS text reminders to reattend between September and December 2012. All patients were included in the analysis even if they had reattended prior to the text being sent.

Control group: patients with the same risk factors as the intervention group who had attended the clinic between September and December 2011 when SMS text reminders were not being used. The control group was matched in number and proportion of STI risk groups to the intervention group and were consecutive and otherwise unselected.

Study size: a minimum of 266 patients were needed in each group to detect a rise in reattendance rate from 20% to 26% using a 5% α-error level and 50% β-error level.8 A change of 5%–6% in reattendance is what would be expected from the results of other studies.5 ,6

Method of evaluation

Reattendance was counted if it was within 4 months of the end of the previous clinical episode.

Analysis

Statistical methods: a two-tailed Fisher's exact test was used to assess the difference in reattendance rates between the SMS text group and the control groups. CIs were calculated using the Wald method.

Results

Participants: a total of 273 people in the SMS text group and 266 people in the control group. Their demographics are shown in the table 1.

Table 1

Demographics, reattendance rates and STI rates at reattendance

Main results: the reattendance rates are shown in the table 1—no differences were found. We looked at why patients eligible for texting were not put on the text recall list in 2012. The reasons were: not offered to patient (44%), patient omitted from recall list (24%), patient recalled by other means (13%), patient declined (11%) and unreliable contact details (8%).

The rate of STIs at reattendance is given in the table 1. The rate of new STIs was the highest in patients recalled after a diagnosis of chlamydia: 9/43 (21%) in the text group and 10/41 (24%) in the control group. The new STI rate in other risk groups was 4/43 (9%) in the text group and 5/51 (10%) in the control group.

Of 273 SMS texts sent, 246 (90%) were confirmed to be delivered by the patients’ mobile provider.

Discussion

Summary

This study shows that SMS text reminders to retest did not increase reattendance rates in high risk patient groups in our clinic population. This is in contrast to previous studies, which demonstrated an increase in reattendance rates following text reminders in patients who had had chlamydia.5 ,6 We found high rates of STIs in patients who did reattend.

Relation to other evidence

Despite the increased risk of testing positive for a STI following a previous infection, the overall rate of repeat testing has been found to be low in other settings.5–7 Our study evaluated the use of SMS text messaging to recall patients to retest. Other methods of recalling patients have also been evaluated elsewhere. Postcard reminders for retesting following a positive chlamydia test have been shown to increase retesting rates from 7.7% to 14.1%.9 Telephone recall of MSM who were diagnosed with a bacterial STI led to a 68.4% retesting rate after 3 months, although the patients were called up to four times.10 A Chlamydia Screening Programme in The Netherlands, in which all chlamydia-positive participants automatically received a test kit after a further 6 months, had a retest rate of 66.3%.11 These studies may therefore suggest that SMS texting is inferior to delivery of home test kits or direct person to person telephoning on multiple occasions as a means of improving retesting rates, although SMS texting is cheaper.12

Limitations

The group of patients we identified was a mixture of several different risk groups.

Only the patients with chlamydia and patients receiving emergency contraception were present in large enough numbers in the analysis to be confident that texting did not make a difference in these subgroups. However, in most cases there was not even a trend towards a benefit from texting and so analysing larger numbers would not likely come to a different conclusion.

This was a retrospective before and after study and in such studies it cannot be proved that any changes that occurred were directly related to the change being measured. However, although we detected no change in the reattendance rate it is possible that any positive effect of SMS text recall could have been negated by other changes in the service system.

Interpretation of results

The contrast in results between this study and previous research may be explained by the relatively high baseline reattendance rate in our 2011 control group, which was 35%. This rate is much higher than the reattendance rates described in the control groups in other published studies, which were 6.3%–21% at baseline and reached 28%–30% after text recall.5 ,6 Therefore, this might suggest that we had reached a saturation point at which a single SMS text reminder was of no added value in a clinic where reattendance rates were already relatively high.

The negative finding in this study might also be explained by the wording of the SMS text message. It may not have been clear that the message was from the sexual health clinic or that it was in response to a previous clinical diagnosis. However, before the patients were put on the text recall list, it was explained to them that such a text would be sent. Texting might have been more effective if it had contained a health message, had been more tailored to each patient or had been linked to better two-way communication or a part of a multi-text strategy.13

Conclusions

This study did not demonstrate an increase in reattendance rates in high risk patients following the introduction of SMS text reminders to retest as compared with verbal advice to reattend. This may be due to the high reattendance rate in the control group at baseline.

Key messages

  • Patients diagnosed with a sexually transmitted infection (STI) are at high risk of further STIs in the future.

  • Previous studies have shown that short message service (SMS) text recall encourages reattendance of patients with chlamydia.

  • We did not see any effect of SMS text recall on the reattendance rate in high risk patients when compared with verbal advice to reattend.

  • The lack of effect of SMS text reminders may be due to the high rate of reattendance already seen in our patients before text reminders were instituted.

Acknowledgments

We would like to acknowledge the efforts of the clinic staff who were key to the success of the SMS texting.

References

View Abstract

Footnotes

  • Handling editor Jackie A Cassell

  • Contributors JB, GB and JM planned the study. JB and GB analysed the data. JB wrote the first draft of the paper. GB, JM and SM contributed to rewriting subsequent drafts.

  • Competing interests None.

  • Ethics approval Ethics committee approval was not required as this study was an evaluation of a service improvement.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

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.