Predicting attendance at child and adolescent clinics

climate in the NHS the numbers seen, whether by a particular service or clinic, or by a particular team or individual, are the subject of ever closer scrutiny. There is pressure to reduce nonattendance rates, although the reasons under lying these may not be under the control of the therapist. The literature suggests areas which might help decide whether a family is likely to continue in therapy. The results of this are in part contradictory, and the overall impression is of the number and diversity of the different factors which might determine whether an appointment is kept or not. Whether looking at socioeconomic (Gould et al 1985; Kazdin, 1990; Beer, 1992; Kazdin & Mazurick, 1994), or demographic factors, at family or parental factors (Cottrell et al 1988; Armbruster & Schwab-Stone, 1994; Kazdin & Mazurick, 1994), or at the pattern of presenting symptoms (Gould et al 1985: Kazdin, 1990: Beer, 1992; Kazdin & Mazurick, 1994) the results are conflicting and no clear conclusions can be drawn. The two most frequently given reasons for dropping out of treatment are that the child does not want to return, and that treatment is no longer required (Gould et al 1985). It is also important to consider what the family might criticise in the psychiatric service. There are issues around convenience, cost, image, Stigmatisation which might prevent them from attending. Once they have been seen, there might still be differing expectations of treatment, per haps in terms of duration, focus or expected outcome (Pekarik, 1985). Beer (1992) also found that staff inexperience, shorter initial interview duration and type of treatment offered, all had implications for continued attendance. This study looks at how well therapists can predict the return of a client, and which reasons tend to be most reliable in helping the therapist to make this decision.

outcome (Pekarik, 1985).Beer (1992) also found that staff inexperience, shorter initial interview duration and type of treatment offered, all had implications for continued attendance.
This study looks at how well therapists can predict the return of a client, and which reasons tend to be most reliable in helping the therapist to make this decision.

The study
This study was undertaken in one of the four consultant led Child and Adolescent Psychiatry teams based at the Brynffynon Clinic in Mid-Glamorgan.Mid-Glamorgan is an area of severe deprivation (Mortimer, 1991).The therapists in this team completed the data form for each separate case seen throughout a three month period from October 1993 to January 1994.They were requested to give patient identification de tails: to state whether they thought the family would attend the next appointment offered: and to state the main reason or reasons they had decided this.

Findings
In this period 80 families were seen.Of these 31 (39%) attended the next appointment offered: 49 (61%) did not.Seventy-five per cent of the predictions about attendance were correct.The therapists predicted that 35 (44%) of the cases would attend; of which 66% were correct predic tions.It was predicted that 45 (56%) of the cases would not attend: of which 82% were correct predictions.
The main reasons for making these predic tions were concerned with the previous record of attendance in 47 of the 80 cases.The reason was given as previous poor attendance for 30 families (25 correct predictions) and previous good attendance for 17 families (12 correct predictions).
Other reasons given included 'clinical assessment of the case' (11 predictions of which 8 were correct); impression of the family's motivation to attend (7 predictions of which 5 were correct); the family's generic social worker had undertaken to ensure that the family would attend (7 predictions of which 6 were correct); and other or no reason given in 9 cases (6 correct).
The therapists' assessment of previous atten dance was a very good predictor of subsequent attendance (x2=12.30P=<0.001).The other rea son given gave some indication of whether the family was likely to attend, but the numbers involved were small.

Comment
The findings from this study might seem selfevident: that those families with a poor atten dance record are unlikely to keep subsequent appointments.
However, what the study does show is that clinicians, without going into meticulous detail about the patient's back ground, socioeconomic status, diagnosis etc. are often able to predict accurately who is likely to attend clinic and who is not.
Why then, if we think that we know this information already, do we not use it?Asking the family to state specifically if they want a further appointment and when they would be able to attend; perhaps telephone prompting (Brockless.1990): or even sending a question naire (Mathai & Markantonakis, 1990) might be effective ways to ensure better attendance.This might be particularly relevant with those families least likely to attend.At least it could mean that the therapist should know whether the family is going to attend or not in time to allocate the appointment to someone else.