Skip navigation

Did not attend

What are rates of nonattendance?
Why do patients not attend?
Can nonattendance rates be reduced?
Comment

These are fateful words for many different healthcare clinics. A patient who fails to attend means that resources allocated to that patient are not used. High rates of nonattendance produce inefficiencies and waste. Three questions arise from this - what sort of rates of nonattendance occur, why do they occur, and is there anything that can be done to reduce nonattendance?

Bandolier sought to find evidence to answer these questions. This was done by referring to a systematic review current up to 1990 [1], and by a search of PubMed for articles on nonattendance looking for surveys and trials performed since about 1990. The main findings from the Bandolier search are shown in the Table, but the search was probably not exhaustive.

Table: Results of Bandolier search for articles related to patient nonattendance at clinics. Studies in grey are controlled trials

Reference Place Setting Number Result
McGlade et al, BMJ 1988 297:1246-8 Belfast First hospital appointment made by GP 269 referrals 15% of all patients failed to attend
Kane, Radiogr Today 1991 57:15-9 Manchester One year audit of outpatient X-ray department 5,323 appointments 5% of all patients failed to attend. Main reason in 51 non-attenders was illness
Lloyd et al, Fam Pract 1993 10:111-7 London ENT and gastroenterology outpatient clinics 1492 first time appointments 26% failed to attend ENT 20% failed to attend gastroenterology
Verbov, JR Soc Med 1992 85:277-8 Liverpool Dermatology outpatients 100 non attenders 28% DNA because of illness, and 33% DNA because of problems related to the appointment
Dockerty, NZMed J 1992 105:147-9 Dunedin, NZ Outpatient appointments at Dunedin hospital over 6 months 37,271 appointments 21% failed to attend
Potamitis et al, JR Soc Med 1994 87:591-3 Birmingham 13 month survey of eye hospital outpatients 5,248 appointments 10% failed to attend. Main reasons were clerical errors and forgetting appointment
Bottomley et al, Clin Exp Dermatol 1994 19:399-400 Leeds New referrals at dermatology outpatients over 12 months Number not available 19% failed to attend. Main reason were forgetting appointment and communication failure
Herrick et al, J Dent 1994 22:307-9 Argyll & Clyde Periodontal clinic Number not available Main reasons for non attendance were forgetting appointment and communication failure
Dini et al, Arch Pediatr Adolesc Med 1995 149:902-5 Atlanta, USA RCT for public health clinic appointments - normal versus computer-generated telephone appointment 517 appointments 68% failed to attend without reminder 48% failed to attend with reminder Statistically significant improvement
King et al, JR Soc Med 1995 88:88-90 Liverpool Ophthalmic outpatients survey over 1 year 43,004 appointments 13% failed to attend.
Ross et al, Genitourin Med 1995 71:393-5 Edinburgh Four clinics surveyed over 1 month Number not available 15% DNA at genitourinary clinic 13% DNA at medical clinic 15% DNA at surgical clinic 14% DNA at dermatology clinic
Komoroski et al, Pediatr Emerg Care 1996 12:87-90 Little Rock, USA RCT for follow-up appointments after emergency department visit, various reminders 253 patients and families 76% failed to attend without reminder 53% failed to attend with simple written reminder 48% failed to attend with written reminder and other interventions Statistically significant improvement
Simmons et al, JR Coll Physicians Lond 1997 31:70-3 Leeds General medical and gastroenterology outpatient clinic, new patients Number not available 38% failed to attend
O'Brien et al, Pediatrics 1998 101:E6 Cleveland, USA RCT of adolescent routine appointments with telephone reminder 703 appointments 57% failed to attend without reminder 35% failed to attend with reminder Statistically significant improvement
Reekie et al, Br Dent J 1998 185:472-4 Manchester Trial of postal reminder versus no reminder in single-handed dental practice 1000 attendances 9% failed to attend without reminder 3% failed to attend with reminder
Stone et al, JR Soc Med 1999 92:114-8 Exeter 6 month prospective survey of plastic surgery outpatient clinic 6,095 appointments 16% failed to attend
Hamilton et al, BMJ 1999 318:1392-5 Exeter RCT of giving referral letter to patients attending outpatients 2,078 referrals 5% failed to attend without copy of letter 5% failed to attend with copy of letter

What are rates of nonattendance?


The review quoted nonattendance rates of 19% to 52%, and found an average nonattendance rate of 43% with a range of 6% to 92%. The more recent information in the Table shows nonattendance rates of between 5% and 38% in UK studies, with 21% nonattenders at clinics in Dunedin, and somewhat higher rates in the USA.

Why do patients not attend?


The studies that have asked this question consistently come up with two major reasons. The first is that patients forgot. The other reason is that clerical errors or communication failures meant that patients did not know they had an appointment.

Can nonattendance rates be reduced?


The systematic review [1] examined randomised trials with quantitative data on the effect of interventions to improve attendance at healthcare appointments. They found 23 trials up to 1990, and the interventions were, in the main, simple telephone or written reminders. For letter and telephone prompts, the reported outputs were as odds ratios, with odds ratios of between 2 and 3.

Three additional randomised trials and one controlled trial of telephone or written prompts were found since the original review (Table, with randomised or controlled trials in grey). They all reduced nonattendance rates and also produced odds ratios of about 2 or 3. Giving patients referred to hospital a copy of the referral letter did not reduce nonattendance rates, though the rate of nonattendance, at 5%, was so low that there was little room for much benefit to be demonstrated.

Odds ratios are not very helpful, and the percentage of nonattenders varies considerably. So the review helpfully generates the number of patients who have to be sent a reminder for one additional patient to attend for their appointment (Figure).

Figure: NNTs for simple telephone or postal reminders depending on the percentage of nonattenders

  • When nonattendance rates are below 10%, 25 have to be sent a reminder for one additional attendance.
  • When nonattendance rates are about 20%, 10 have to be sent a reminder for one additional attendance.
  • When nonattendance rates are about 35%, 6 have to be sent a reminder for one additional attendance.
  • When nonattendance rates are above 50%, 5 have to be sent a reminder for one additional attendance.

Comment


Bandolier has not done a complete review with full meta-analysis and economic assessment here, partly because of time and partly because it isn't needed. We know that nonattendance rates are variable, if often too high. We know that simple interventions are effective, and similarly effective, across a range of nonattendance rates. We know that every clinic has its own idiosyncrasies which makes a nonsense of generalities about economic assessment. A problem may be that all the randomised trials are American, and have high nonattendance rates. This may make generalisation problematical.

The one generality that is of use is the NNT calculation. Any clinic could audit its nonattendance rate and judge the cost and consequences of instituting simple postal or telephone reminders.

Reference:

  1. WM Macharia et al. An overview of interventions to improve compliance with appointment keeping for medical services. JAMA 1992 267: 1813-1817.
previous or next story in this issue