Results

The proportion of non-attenders in the Asian and English groups was compared by χ² tests. Significant differences were found between the new Asian and new English groups, that is, 50%v 32% (χ² = 54.5, p<0.001) and between the follow up Asian and English groups—50% v 33% (χ² = 15.5, p<0.001) (table 1) When the new English patients who failed to attend were compared with English follow up patients who failed to attend, and a similar exercise carried out for the Asians, there was no statistically significant difference (χ²=0.057, p<0.81 and χ²=0.004, p<0.95 respectively). The majority who did not attend among the Asians were Muslim.

Generally, non-attendance was higher among Asian boys than Asian girls, particularly among the Muslims (table 3). Twenty six out of 49 Sikh boys and 25 out of 41 Hindu boys failed to attend. Statistical analysis showed that there was significantly more failure to attend among Muslim boys than among Muslim girls (χ²=46.82, p<0.001). No such sex difference was statistically significant for Hindus, Sikhs, or the English.

Table 2 shows that failure to attend follow up appointments from the wards among Asians are 63% compared with 41% among the English, and 46% for Asians compared with 31% for English when outpatient department generated follow ups are analysed.

The results of comparison between the number of patients who did not attend after inpatient stay with those who did not attend subsequent to their outpatients appointment are: Asian ward follow up failure to attend v outpatient department failure to attend (χ²=64.05, p<0.001), English ward follow up failure to attend v outpatient department failure to attend (χ²=71.61, p<0.001), Asian ward failure to attendv English ward failure to attend (χ²=141.1, p<0.001). Comparison of all Asian outpatient department failure to attend v English outpatient department failure to attend (χ²=177.3, p<0.001).

Table 4 shows the main reason for not attending outpatient departments in 120 children studied. Almost a third of Muslims failed to attend because of religious holidays or being abroad. Although the Bangladeshis form the smallest proportion of the Muslim community in Walsall, the largest being Mirpuri Pakistanis, there were nine out of a total of 20 Bangladeshi children who did not attend because of communication difficulties.

Figure 1 shows a peak of non-attendance for Muslims in February, May, August, and December, with the English having the main peak in August. When monthly comparisons are made using the χ² tests and Bonferroni correction,8 failure to attend among Muslims was significantly higher during February, May, August, and December (p<0.05). For non-Muslim Asians the percentage of non-attendance was significantly higher during August when this month was compared with non-attendance from other months (p<0.05). There were no such significant statistical differences among the English.

Discussion

The months of February and May 1995 coincided with the fasting month (Ramadan) and the time for pilgrimage to Mecca (Haj) respectively. During these months, particularly the former, there was a peak of failure to attend among Muslims, which was also evident in December (fig 1), a period during which many Muslims go for Umra (a lesser form of pilgrimage) and visit relatives abroad, especially those from the Indian subcontinent. As Muslims follow a lunar calendar, Ramadan begins and ends 10 days earlier in succeeding years. The August peak among all groups is due to the fact that schools are closed for the summer holidays in this month.

The first four reasons for failure to attend listed in table 4constitute 60% of patients not attending. The other reasons given in table 4, in our experience, would also apply to those failing to attend among the English.

There is a striking difference in non-attendance when follow up Asian failure to attend (63%) from the ward is compared with the English failure to attend (41%) (table 2). The main reason is probable difficulty in communication by nurses on ward discharge and the fact that the child was better when the appointment was due. This communication aspect was not studied specifically for these non-attendance cases, although in the overall sample there were 20/120 such patients (table 4). Interestingly, there were a significant number of ward generated English outpatient patients failing to attend, suggesting that it is possible that unnecessary return appointments are given to patients whose parents perhaps feel that their child has recovered from their acute episode and that there is no need for reattendance. Visits to a doctor when one feels well are not part of the culture of people from Southeast Asia. This philosophy is continued in the UK, particularly by the first and second generation of Asians.

The Muslim religion encourages education for girls,9 but in reality this is less of a priority than education for boys and so it is possible that more girls than boys kept their appointments because their parents were more willing for them to miss school (table 3). The fact that Muslims are a predominant group in our study means it is possible that this has resulted in the overall Asian non-attendance for boys being significantly greater than that for girls (p<0.001). No such differences were observed among boys and girls in the Hindu, Sikh, or the English groups.

Although Asian women generally prefer to see a female doctor (particularly for gynaecology), this is not as critical for paediatrics. Indeed, there was no significant difference in non-attendance for appointments with the woman consultant, when compared with either of the male paediatricians and, furthermore, there was no statistical difference in failure to attend when the two English consultants were compared separately with the Asian consultant. As Asian children are often threatened with injections at home for bad behaviour,10 it is possible that some children, particularly those between the ages of 5–10 years, refuse to be brought to the outpatient department.

Our study would appear to support the findings of others,11 that the weather has little effect on failure to attend. Summer months (April to September) and winter months (October to March) showed very little difference in non-attendance for the Asians (50% in summer) and the English (52% in summer) (fig 1). This is worth emphasising, as during the summer months one would expect greater non-attendances with holiday periods, etc—perhaps this is offset by the fact that the adverse winter weather could also increase failure to attend.

Failure to attend the hospital for outpatient appointments can increase with the cost of transport.12 Most of our Asian patients live in the Pleck, Palfrey, Chuckery, and Darlaston areas of Walsall, which are nearest the hospital and therefore this reason can be largely discounted.

Although Andrews et al showed that non-attendance was detrimental to a child’s health, the longer waiting lists could result in the child being better, parents forgetting, the child being admitted into hospital, or being seen privately.7

The average waiting list for paediatrics at the Manor Hospital, Walsall, during 1995 was about five weeks, which, although not ideal, is well within the Patients’ Charter guidelines.13 In spite of this relatively short waiting list, the non-attendance among new Asian patients was 50%.

Our total paediatric non-attendance for follow up appointments was 35%, which is comparable with those cited by Andrews et al. However, our 33% of non-attendance for all new patients is higher than their 25%.7 Suprisingly, when we referred to the Hospital Aggregate Returns Report of 1994–5 (the data for which is gathered electronically), the overall non-attendance rate for Walsall was 24.4%.14 I find this hard to accept as our monthly collection was done regularly by myself. Other neighbouring hospitals, Sandwell, Dudley, and Wolverhampton scored 29.5%, 22.3%, and 20.9% non-attendance respectively.

One of the criticisms of our study is with regard to the fact that if all the patients had been seen by the consultant, then there may have been a lesser number of follow ups requested, although the reason for follow up or otherwise would differ among consultants.

To decrease Asian non-attendance, some knowledge of their culture and religion is essential. The following recommendations may be helpful:

• Better communication, including the use of link workers for patients from all communities, particularly those from Bangladesh, many of whom only speak Sylethi and not Hindi which is the main spoken Asian language in the UK.

• Better signposting in hospitals.

• Outpatient clerks should have a list of dates for the fasting month and religious festivals, including pilgrimage. As Friday is a religious day for Muslims, with prayers at around midday, appointments either late in the morning or early afternoon may not be kept. Similar information can be obtained for other ethnic minorities from a cultural awareness group set up to help hospital administration with such issues. Patients should be specifically asked whether an appointment clashes with their travel abroad or religious holidays.

• Midwives should be aware that Muslim ‘nursing’ mothers may stay at home for six weeks as part of their culture, and perhaps not attend outpatient appointments. Furthermore, a grieving Muslim wife may not leave her home for up to four months after her husband’s death. Cultural awareness groups could communicate with local mosques, temples, etc and advise elders to impress on the community the importance of keeping appointments, and also bring to their attention communication services in ethnic languages provided by the hospitals (if this is the case).

• A study could be undertaken to compare the cost of consultant domiciliary visits with that of non-attendance, particularly in the case of those Asians who have the greatest difficulty in communication, for example, the Bangladeshis.