Attendance rate in an alcohol problem clinic

The high rate of non-attendance for firstappointments at psychiatric clinics iswasteful of health service resources. Previous studies have demonstrated reductions in the non-attendance rates when referred patients were asked to contact the clinic to confirm that they wished to be seen before an appointment was sent. The effect of introducing such a process to a community based alcohol service was investigated. Results showed that although there was a reduction in the non-attendance rate, the overall effect was to discourage patients from attending the clinic, resulting in a large reduction in the proportion of all referrals seen. Alternative approaches are discussed.

Non-attendance of new out-patients at psychia tric clinics is a common problem which is wasteful of clinicians' time and other resources. The rate of non-attendance for first appoint ments has been shown to be around 30% in a general psychiatric clinic (Skuse, 1975). between 15-50% at child psychiatry clinics (Plunkett, 1984;Gould et al 1985;Cottrell et al 1988) and between 12-52% in alcohol treatment clinics (Addenbrooke & Rathod, 1990: Goldbeck. 1993. One study in a general adult psychiatric clinic showed that patients were more likely to attend if they had a previous psychiatric history, the appointment was held at a health clinic rather than in a psychiatric hospital, and patients received a personal letter rather than an im personal, standard appointment card (Hillis & Alexander, 1990). In child psychiatry clinics the attendance rate was not increased by sending families a hand-written appointment letter (Davison, 1992) but was increased when contact was made with families by telephone or post to gather family data prior to the offer of the appointment (Mathai & Markantonakis, 1990). A recent study showed that attendance rates could be increased in a general adult psychiatry clinic by sending patients a reminder within a few days of their appointment (Rusius, 1995).
Asking patients to confirm in advance whether or not they require an appointment has been shown to reduce significantly the non-atten dance rate in two studies. Baggaley (1993) showed a reduction in the non-attendance rate from 28% to 3% in general psychiatry patients but reported that the proportion of all referrals seen fell from 72% to 61%, presumably because some patients who would have attended failed to contact the clinic for an appointment. A sig nificant reduction in the non-attendance rate from 50% to 34.5% was also found following the introduction of a similar system in an alcohol clinic (Goldbeck, 1993). Again there was a fall from 50% to 42% in the overall proportion of total referrals seen.
In this study, an attempt was made to increase the attendance rate for new referrals to a community-based alcohol service by employing a new appointments system similar to those described in these two studies (Baggaley, 1993;Goldbeck, 1993). Clinics were already held in health clinics rather than hospitals and personal letters of invitation were already employed (Hills & Alexander, 1990).

The study
This was a retrospective study of new alcohol referrals to a community drug and alcohol service serving Mid Glamorgan, population 544 300. Prior to January 1994 all referred patients were routinely sent an appointment and those who failed to attend were sent a second and sometimes a third appointment. A new system was adopted which involved writing to all referred patients asking them to contact the clinic by telephone to confirm whether or not they required an appointment. Those who failed to contact the clinic within 8 weeks were not sent an appointment and the referrer was advised that the patient had not responded.
The study involved analysis of case records of new referrals for alcohol problems, excluding self-referrals, referrals from probation officers or solicitors, and domiciliary or hospital liaison requests. Demographic characteristics, atten dance rates and intervals between referral and appointment dates were collected for all relevant patients. The first three months of the new system were excluded to avoid possible complica tions arising from the change-over period and any teething problems with the system. To reduce the possible confounding effect of seasonal variation in attendance rates, referrals from 1 April 1993 were compared with referrals from 1 April 1994.

Findings
There were 120 alcohol referrals between 1 April and 31 December 1993, of which 80 met inclusion criteria for the study, and 125 referrals between 1 April and 3 October 1994 of which 90 met inclusion criteria for the study.
There were no significant differences in demo graphic characteristics between the two groups: the mean age was 39.8 years and the male to female ratio was approximately 2:1. There were similar mean intervals between the date on which the referral was received by the service and the date of the appointment in both groups (44 and 47 days). Those who attended their appointments waited longer for that appoint ment than those who did not attend under both the old (49 u. 37 days) and new systems (49 v. 42 days). Under the old system 48 of 80 patients attended their first appointment, a non-atten dance rate of 40%. Under the new system 42 of 90 patients did not ring to confirm that they wanted an appointment and one person rang to indicate that they did not wish to be seen. Of the 47 patients who rang to confirm their wish to be seen, 31 attended their first appointment, a nonattendance rate of 33%. The difference failed to reach statistical significance. Under the old system 60% of all referrals were seen at their first appointment, compared with 34% under the new system (P<0.001).

Comment
Our findings are consistent with those of the two previous reports (Baggaley, 1993;Goldbeck, 1993) which suggested that asking patients to make contact with a clinic to confirm that they want an appointment can reduce non-atten dance rates in those who respond. This is unsurprising as those with sufficient motivation and organisational ability to reply are, at face value, more likely to attend.
If the data from the current study are pooled with those of Baggaley (1993) and Goldbeck (1993), to make an aggregated total sample size of 608 referrals, then while there is a significant fall in the rate of non-attendance from 42% to 27% (P<0.01), this has to be set against a significant overall reduction in the proportion of all referrals assessed, from 48% to 38% (P<0.001).
The low rate of contacting the service for appointments may be explained by a number of factors, some of particular relevance to patients with psychiatric and alcohol problems. Denial and lack of insight, the stigma associated with seeking help, fluctuating motivation to address one's problems and low self-esteem may all discourage the patient from contacting the clinic. Limited availability of a confidential telephone line may also have been of significance in the socio-economically deprived population under study.
These new appointment systems clearly pro duce potential savings in professional's time but at the expense of accessibility to patients who may be those with most need. In the current study, as many as 43% of those who would have attended under the old system were not seen. A large proportion of patients were thus deprived of the opportunity for assessment and access to treatment.
Alternative options to the system described in this paper merit further investigation to mini mise wasted staff time while retaining and maximising accessibility. Clear communication with and education of referrers might help to reduce inappropriate and obviously unwilling referrals. Moving clinics away from psychiatric hospitals may be impractical in some services, but sending personal letters rather than cards (Hillis & Alexander, 1990) and following up all appointment letters with reminders (Rusius, 1995) could relatively easily be introduced into most administrative systems. Collecting preassessment information by telephone or postal questionnaire might be investigated for adult patients as it has been shown to be successful with families (Mathai & Markantonakis, 1990). In addition, an alternative to telephone or postal contact might be to book all referred patients in to 15-minute, pre-interview assessments with a nurse. At this meeting personal contact could be established, anxieties allayed, routine data and rating scales completed and laboratory screening tests carried out, so that all these data could be available at the time of the assessment interview proper. With only 15 minutes of staff time wasted per non-attender, and the investment made by the patient in the 15 minutes, we would expect that the second, full assessment interview would invariably be an appointment to be kept. ADDENBROOKE, W. M. & RATHOD, N. N. (1990)