Use of mobile telephone short message service as a reminder: the effect on patient attendance

Abstract

Background: Nonattendance for hospital outpatient appointments is a major burden on healthcare systems and the estimated financial loss can be high. Various reasons for patients not attending have been reported, including the patient/carer forgetting the appointment and confusion over the date and time of the appointment. Various reminder systems have been trialled across a variety of clinical settings. More recently, short message service (SMS) text messaging has emerged as a viable approach for delivering reminders to outpatients at a relatively low cost. Aims and objectives: To evaluate the effect of appointment reminders, sent as SMS text messages to patients’ mobile telephones, on attendance at outpatient clinics at the ITS Centre for Dental Studies and Research (ITS-CDSR), Muradnagar, Ghaziabad, Uttar Pradesh, India. Materials and methods: The study was conducted at ITS-CDSR in the Departments of Prosthodontics, Endodontics, Orthodontics and Paedodontics. Patients attending these departments for a period of 4 months and those who had provided a contact mobile number were sent an SMS reminder. Results: In this study, 206 subjects (male, 124; female, 82) participated, 96 (male, 57; female, 39) of whom were in the test group and 110 (male, 67; female, 43) of whom were in the control group. The rate of attendance on time was found to be significantly higher in the test group (79.2%) than in the control group (35.5%). Conclusion: The study results indicate that sending appointment reminders as text messages to patients is an effective strategy to reduce nonattendance rates.

INTRODUCTION

India is the world’s fastest growing wireless market, with 752 million mobile phone subscribers as of February 2011. It is also the second largest telecommunication network in the world in terms of number of wireless connections after China¹.

Nonattendance for hospital outpatient appointments is a major burden on healthcare systems. It is an intractable problem. Many reasons for patients not attending have been reported, including the patient/caregiver forgetting the appointment and confusion over the date and time of the appointment².

Previously, various reminders have been assessed, with varying degrees of success in reducing broken appointments, including the use of automated and personalised phone calls, but they are costly and time consuming³. Therefore, efforts to search for cost-effective interventions is gaining importance in practice management.

Short message service (SMS) text messaging has emerged as a viable approach. SMS facilities on mobile phones were introduced in the early 1990s and now represent one of the most widely used methods of communication.

Text messaging offers a number of benefits over other techniques. It is a cheaper and more efficient method. SMS has the capability for large batches of individualised group messages to be sent at one time. SMS messages have a number of characteristics that make them appropriate for use in a healthcare setting, including direct patient communication, privacy, confidentiality, swift delivery of messages and receipt of responses, and convenience for health providers and patients⁴.

To date, there has been a paucity of research in relation to the use of mobile phone SMS/text message reminders for outpatient hospital appointments.

Thus, the aim of this study was to evaluate the efficacy of appointment reminders, sent as SMS text messages to patients’ mobile telephones, in comparison with no reminders given to patients, at outpatient clinics at the ITS Centre for Dental Studies and Research (ITS-CDSR), Muradnagar, Ghaziabad, Uttar Pradesh, India.

MATERIALS AND METHODS

Study design and setting

This study was performed to test the effect of SMS reminders on patient attendance at ITS-CDSR, comparing a test group with a control group.

The patients included in this study were selected from those who received dental and preventive care at an institution involving four departments, namely Prosthodontics, Orthodontics, Endodontics and Paedodontics.

It was conducted for a duration of approximately 4 months, i.e. from September 2010 to December 2010, after seeking approval from the Institutional Review Board of the institute. Written permission was obtained from the concerned heads of departments who were willing to cooperate with the study. Consent was obtained from postgraduate students who agreed to give their patients’ contact numbers.

The patients were eligible for inclusion if:

• •They were scheduled to attend the selected four departments in the college from September 2010 to December 2010
• •They had given their contact number to the dentists included in the study
• •They had a mobile phone with text messaging services.

Patients who did not provide contact details were excluded from the study.

The four departments were selected on the basis of their large patient volume and relatively high long-term failure to attend rates.

The selection of the subjects was performed randomly. Randomisation was carried out on a departmental basis. Because the randomisation scheme was based on appointments rather than on patients, some of the patients who had more than one single scheduled visit were included in both the test and control groups at some time over the 4-month study period.

A total of 206 subjects participated in the study, 96 of whom were in the test group and 110 of whom were in the control group (Table 1).

Table 1.

Allocation of subjects into test and control groups

Section	Name of department	Test group	Control group
1	Prosthodontics	23	22
2	Orthodontics	10	14
3	Endodontics	49	56
4	Paedodontics	14	18
	Total	96	110

The subjects selected were divided randomly into the test and control groups. Subjects in the test group were sent an SMS reminder, whereas those in the control group were not sent an SMS reminder, or any other type of reminder.

Interventions

The appointments selected for the study were identified on two working days before the scheduled visit. The reminder was sent to the patient or guardian for whom the message was intended.

An SMS reminder was sent 24 h prior to the appointment and also on the day of the appointment. Information for the reminder was extracted from the department scheduling system.

The outcome of the intervention and the identification of the recipient of the message were recorded on the schedule.

The reminder included the dentist’s name, date, time and location of the appointment. The success of delivery of the SMS was assumed when there were an indication showing that the message had been sent on the mobile phone (Figure 1).

Figure 1.

Mobile message received report.

When appointment reminders were being scheduled, 22 patients stated that they did not understand the English language. Reminders for these subjects were sent in the local language on the mobile phone. In addition, a picture message of the institution was sent to seven patients.

Outcome measures

A single outcome measure, the ‘attendance rate’, which is the percentage of scheduled appointments being attended on a scheduled date and time, was used.

Statistical analysis

The data obtained were subjected to statistical analysis using SPSS version 18 (SPSS Inc., Chicago, IL, USA). MS Excel and MS Word were used to generate tables and graphs.

Statistical tests were performed using the chi-squared test, Fisher’s exact test and odds ratio (when required) to test the significant differences in proportions.

RESULTS

The overall distribution of the study population in the test and control groups is shown in Figure 2. Of the total of 206 patients, 96 patients were in the test group (male, 57; female, 39) and 110 patients were in the control group (male, 67; female, 43).

Figure 2.

Overall distribution of study population in test and control groups.

Figure 3 shows the distribution of the 206 patients into 60% (n = 124) males and 40% (n = 82) females.

Figure 3.

Graphical representation showing gender distribution.

Figure 4 shows that a greater number of patients reported from the Endodontics Department, followed by the Prosthodontics, Paedodontics and Orthodontics Departments, respectively.

Figure 4.

Department-wise distribution of study population in test and control groups.

The department-wise response rates with 95% confidence intervals (95% CI) were 55.56% (40–70.36%), 45.83% (25.5–67.18%), 51.43% (41.47–61.30%) and 78.12% (60.02–97.72%) in the Prosthodontics, Orthodontics, Endodontics and Paedodontics Departments, respectively (Table 2).

Table 2.

Department-wise attendance rates and failure to attend in test and control groups

Department name	Group	Patient reported
		On date and time	Not on time	Total	P value
Prosthodontics	Test group				0.000
	Count	21	2	23
	% within group	91.30	8.70	100
	Control group
	Count	4	18	22
	% within group	18.10	81.80	100
Orthodontics	Test group				0.22
	Count	6	4	10
	% within group	60	40	100
	Control group
	Count	5	9	14
	% within group	35.70	64.30	100
Endodontics	Test group				0.000
	Count	37	12	49
	% within group	75.50	24.50	100
	Control group
	Count	17	39	56
	% within group	30.40	69.60	100
Paedodontics	Test group				0.32
	Count	12	2	14
	% within group	85.70	14.20	100
	Control group
	Count	13	5	18
	% within group	72.20	27.80	100

A significant P value of 0.000 was observed in the Prosthodontics and Endodontics Departments, which reported higher attendance rates (Table 3).

Table 3.

Chi-squared test

Department	Value	df	Asymptotic significance (two-sided)	Exact significance (two-sided)	Exact significance (one-sided)
Prosthodontics
Pearson chi-squared	24.350*	1	0.000
Continuity correction†	21.478	1	0.000
Likelihood ratio	27.374	1	0.000
Fisher’s exact test				0.000	0.000
Linear by linear association	23.809	1	0.000
Number of valid cases	45
Orthodontics
Pearson chi-squared	1.386‡	1	0.239
Continuity correction†	0.58	1	0.446
Likelihood ratio	1.395	1	0.238
Fisher’s exact test				0.408	0.233
Linear by linear association	1.328	1	0.249
Number of valid cases	24
Endodontics
Pearson chi-squared	21.330§	1	0.000
Continuity correction†	19.36	1	0.000
Likelihood ratio	22.17	1	0.000
Fisher’s exact test				0.000	0.000
Linear by linear association	21.127	1	0.000
Number of valid cases	105
Paedodontics
Pearson chi-squared	0.839¶	1	0.36
Continuity correction†	0.235	1	0.628
Likelihood ratio	0.867	1	0.352
Fisher’s exact test				0.426	0.318
Linear by linear association	0.813	1	0.367
Number of valid cases	32

^*0 cells (0%) have an expected count <5. The minimum expected count is 9.78.

^†Computed only for a 2 × 2 table.

^‡1 cell (25.0%) has an expected count <5. The minimum expected count is 4.58.

^§0 cells (0%) have an expected count <5. The minimum expected count is 23.80.

^¶0 cells (0%) have an expected count <5. The minimum expected count is 3.06.

Table 4 shows that the response rate in the test group is 79.20%, whereas, in the control group, the response rate is 35.50%. The total overall response rate with 95% CI is 55.83% (48.76–62.72%). In Table 5, a P value of 0.001, showing a statistically significant difference between the test and control groups, was observed. In Table 5, the odds ratio was calculated, and it was found that there was a 6.9 times greater chance that the patient would report on time in the test group relative to the control group, with a 95% CI of 3.69–12.97 and a degree of freedom of unity (Table 6).

Table 4.

Comparison of attendance rates on scheduled date and time in test and control groups

	Patient response
	Visit on scheduled day	Did not visit on scheduled day	Total
Test group	76	20	96
% within group	79.20	20.80	100
Control group	39	71	110
% within group	35.50	64.50	100
Total	115	91	206
% within group	55.80	44.20	100

Table 5.

Risk estimate

	95% confidence interval
	Value	Lower	Upper
Odds ratio	6.918	3.689	12.972
Reported (1/2)
For cohort group = 1	3.007	1.998	4.525
For cohort group = 2	0.435	0.329	0.574
Number of valid cases	206

Table 6.

Chi-squared test

Test	Value	df	Asymptotic significance (two-sided)	Exact significance (two-sided)	Exact significance (one-sided)
Pearson chi-squared	39.719*	1		0.000
Continuity correction†	37.966	1	0.000	0.000
Likelihood ratio	41.473	1	0.000	0.000
Fisher’s exact test			0.000	0.000	0.000
Linear by linear association	39.526	1
Number of valid cases	206		0.000

^*0 cells (0.0%) have an expected count <5. The minimum expected count is 42.41.

^†Computed only for 2 × 2 table.

Table 7 shows the differences in rates of attendance on the specified date and time between different departments. In the test group, 73.2% of patients reported on the specified date and time in the Prosthodontics Department, whereas only 13.5% of patients reported on the specified date and time in the Paedodontics Department.

Table 7.

Differences in rates of attendance on time between departments

Section	Name of department	Patient on time (%)
1	Prosthodontics	73.20
2	Orthodontics	24.30
3	Endodontics	45.10
4	Paedodontics	13.50

DISCUSSION

The present study was designed with the aim of evaluating the impact of appointment reminders, sent as SMS text messages to patients’ mobile telephones, in comparison with no reminders, at outpatient clinics at ITS-CDSR, Muradnagar, Ghaziabad, Uttar Pradesh, India. The use of SMS technology appears not to have been widely investigated as a means of improving patient rates. This study showed that a text messaging reminder was effective in improving attendance in primary care compared with a no-reminder control. It was also found to be more cost-effective than mobile phone call reminders. This result may have a significant impact on the choice of reminder system in primary care². Other forms of reminders to patients, such as voice messages and personalised phone calls, are very difficult to implement. The search for a cheaper reminder system was necessary, and an initial step has been taken to evaluate the efficacy of this new and emerging technique in increasing outpatient attendance.

In the present study, the rate of attendance on time was higher in the test group (79.2%) than in the control group (35.5%), which showed a statistically significant difference. The reminders sent to patients attending the Prosthodontics Department were highly significant, as there was a difference of about 73.2% between the test and control groups, whereas little difference, only 13.5% between the test and control groups, was seen in patients attending the Paedodontics Department. This may be a result of the fact that, in the Paedodontics Department, reminders were sent to the caregiver.

The limitations of this study are that we do not know whether patients came for their appointments because of their own prior decisions, or because we sent them an SMS text message. Moreover, economic aspects need to be considered in mobile phone usage, including messaging charges and schemes. To date, little has been published on the use of SMS text messaging to send appointment reminders to patients’ mobile telephones.

Few studies have reported on this topic in the dental literature. From an internet search, a study by Nelson⁵ compared a voice message system with text messaging. This dental study showed voice messages to be more effective than text messaging, in contrast with the present study, where an SMS reminder was found to be effective in increasing outpatient attendance. This may be because voice reminders are more interactive than text messages, which can be easily ignored.

In the present study, the results were found to be in favour of sending SMS reminders. Other studies have also reported similar results. For example, Downer et al.⁶ found that the overall failure to attend rate was significantly lower in the trial group than in the control group (14.2%vs. 23.4%). This study also suggested that SMS is a suitable means of improving patient attendance.

Nonattendance rates were 38% lower in patients who received an SMS reminder than in those who did not receive a reminder in a study conducted by Koshy et al⁷. Chen et al.⁴ also found that the attendance rates were significantly higher in SMS and telephone groups than in the control group. However, there was no difference between the SMS group and the telephone group. An SMS reminder proved to be more cost-effective than a telephone reminder. da Costa et al.⁸ concluded that sending appointment reminders as text messages to patients’ cell phones is an effective strategy to reduce nonattendance rates. Leong² found that the attendance rates of control, text messaging and mobile phone reminders groups were 48.1%, 59.0% and 59.6%, respectively. The attendance rate of the text messaging group was significantly higher than that of the control group, but there was no difference between the text messaging group and the mobile phone group.

The present investigation is in agreement with most of these studies; thus, SMS reminders are simpler and more easily applicable in the present electronic usage system.

Further studies

Text message reminders are considered to be cheaper than other reminder methods, such as personalised phone calls and voice reminders. Moreover, the dentist may not wish to call the patient personally, as the patient may wish to have discussions with the dentist on the phone, thus taking up the dentist’s time. Phone calls and voice reminders are more difficult to ignore than SMS text messages, which can easily be ignored by the patient. The main advantage of SMS text messaging is that it is cheaper, as SMS reminders require minimal investment in IT infrastructure in order to send automated SMS text messages using appropriate systems. Once the system is in place, the cost of running the service increases slightly as the number of SMS reminders increases.

The reasons for follow-up may have an impact on attendance rates and, possibly, on the effectiveness of different reminder systems. Further studies looking at the effectiveness of text messaging reminders in different types of follow-up are warranted.

CONCLUSION

The study results indicate that the sending of appointment reminders as text messages to patients is an effective strategy to reduce nonattendance rates. Nonattendance at outpatient clinics is a common intractable problem, found across a wide variety of healthcare disciplines, resulting in the underutilisation of resources and prolonged waiting schedules. For the majority of dental staff, the sending of text messages, leading to an improvement of nonattendance at outpatient clinics, is definitely beneficial. Further research of this method in all dental healthcare delivery systems should be carried out.

Acknowledgements

The authors wish to thank the Departments of Prosthodontics, Orthodontics, Endodontics and Paedodontics for their co-operation.

Conflicts of interest

The authors declare that they have no conflicts of interest.