O Adebami, J Owa, G Oyedeji
attrition rate, fetal malnutrition, follow-up, nigeria
O Adebami, J Owa, G Oyedeji. Attrition Rate of Follow up Attendance in a Western Nigerian Fetal Malnutrition Study. The Internet Journal of Nutrition and Wellness. 2007 Volume 5 Number 1.
Attrition rate at the 4 weeks follow up was higher among the mothers who had no antenatal care compared with those who did (χ2=11.3, p =0.001), among teenage than older mothers (χ2=3.1, p =0.08), among primiparous than multiparous mothers (χ2=2.5, p =0.1) and among lower than higher social class mothers (χ2p= 0.005). Attrition was also higher among mothers of female than male sex babies (χ2=4.5, p=0.03), mothers of babies without fetal malnutrition than those with FM (p=0.5), among clinically well than ill babies requiring admissions (χ2=39.7, p=0.000).
To reduce the frequency of attrition in a research study, better education of subjects and enlistment of the services of social worker/health visitors should be undertaken. Researchers in the developing nations should make generous allowance for attrition in calculating the number of the subjects to be recruited. This may be higher than what obtains in the developed nations. Research work should be made relevant to subjects need and planned in a way that subject participation would not overstretched their resources. Improvement in infrastructure including transportation and road network would also assist research activities.
Attrition is a major factor for consideration when carrying out research work involving patient follow up or cohort design. The number estimated as would be lost to attrition is usually made allowance for in calculating the number of the subjects needed for the study. Patronage of modern health facilities in the developing countries is low. This has been attributed to long distances to the facilities, lack or high cost of drugs and services, long waiting time, gender preferences, poor quality of care and preferences for traditional healers.2,3,4,5 Also, poor transportation system, incessant and abrupt stoppage of services through strikes at very critical period of care, ignorance, increasing poverty among others were said to have hindered health facilities utilization in some communities.2,3,4,5,6 At times, subjects who have given informed consent change their mind and bolt away from the study. To our knowledge there has been no study in Nigeria reporting the attendance rates of subjects to follow up clinic in a research study situation. The aim of the present study was therefore to examine the role of maternal and child factors in attrition rates in follow up attendance in Nigeria. The study was conducted at the maternity and neonatal units of the Wesley Guild Hospital (WGH), Ilesa, Southwestern Nigeria. This is the main referral health institution providing general and specialist Paediatrics and maternity services to the semi-urban Ijesa community of Osun State in Southwestern Nigeria.
During the period between January and August 2001, Four hundred and seventy three consecutive, singleton, live birth, term (37-42 weeks gestation) babies delivered at the Wesley Guild Hospital, Ilesa, Nigeria were studied for the prevalence and other details of fetal malnutrition. The details of methodology of the study and the findings ensuing from it are as reported in another paper.7 It was planned at the time of study of each patient that he/she would be followed up for 4 weeks or more (if the delivery was complicated). This would involve clinic attendances at 1 week and 4 weeks of age at the minimum. The details were fully explained to the parents and their consents were obtained. The mother/parents gave a general consent including that of attendance for follow up.
The follow up attendance rates of the 473 babies at the 4th week of life were recorded and compared in relation to maternal age, parity and socioeconomic classes and attendance of antenatal clinic as well as babies sex and coexistence in them of fetal malnutrition and other clinical problems. Babies were classified into those with fetal malnutrition (FM) and those without FM.
Of the 473 expected for follow up, 315 (66.6) and 249 (52.6%) attended the follow up clinic at the first and fourth weeks respectively. The two hundred and twenty four (47.4%) lost to follow up at the 4th week therefore formed the subjects studied with respect to the maternal and baby factors listed above under methods.
Of the 473 total subjects recruited, 246 [52.0%] males and 227 [48.0%] females giving a male: female ratio of 1.08:1. Male and female clinic attendance were 141 (56.6%) and 108 (43.4%) respectively meaning that 105 (46.9%) of the 246 males compared with 119 (53.1%) of the 227 females defaulted. Therefore, significantly higher proportions of female sex were lost to attrition than males (χ2=4.5, p=0.03).
Maternal factors:Places of antenatal care and Social Class
On the whole, 394 (83.3%) of the mothers had antenatal care in various health units. Altogether, 173 (43.9%) of the 394 who had antenatal care compared with 51 (64.6%) of 79 who had no antenatal care defaulted at the 4 week follow up (χ2=11.3, p =0.001).
The ages of mothers ranged between 15 and 45 years. There were 23 (4.9%) mothers aged 19 or less, 363 (76.7%) mothers aged 20-34 and 87 (18.4%) aged 35 and above. Fifteen (65.2%) of the 23 teenage (age ? 19 years) mothers compared with 209 (46.4%) of the 450 mothers who were over 19 years defaulted (χ2=3.1, p =0.08). Parity of the mothers ranged between 1 and 9. One hundred and fifty six (33.0%) were primiparous and 317 (67.0%) were multiparous. Higher proportions of primiparous mothers though not statistically significant were lost to attrition. Thus, 82 (52.6%) of 156 primiparous mothers versus 142 (44.8%) of 317 multiparous mothers defaulted (χ2=2.5, p =0.1).
Also, there were 34 mothers in social class I, 95 in class II, 123 in class III, 145 in class V and 76 in class V. One hundred and twenty (54.3%) of the babies of the 221 mothers in lower social classes IV and V were lost to attrition compared with 104 (41.3%) of the babies of the 252 mothers in social class I to III. (?2 p= 0.005). Therefore, 120 (53.6%) of the 224 babies lost to attrition were delivered by 221 mothers in lower social classes IV and V.
Associated Neonatal Problems
Overall, 89 (18.8%) of the 473 babies had fetal malnutrition. There was no statistically significant difference in the clinic attendance between babies with FM and those without, though higher proportions of babies without FM were lost to attrition. Thus, 39 (43.8%) among the 89 babies with FM defaulted compared with 185 (48.2%) among 384 babies without FM (p=0.5).
One hundred and ten (23.3%) of the 473 babies were admitted for different adverse clinical conditions. Twenty nine (26.4%) of the 110 babies who had clinical problems and were admitted compared with 220 (60.6%) of the 363 babies with no clinical problems and hence not admitted were lost to attrition ( = 39.7, p=0.000). None of the three babies with severe birth asphyxia who had sequelae of hypertonia, hyper-reflexia, poor sucking and seizures defaulted from the clinic during the period.
The overall attrition level in the present study was 47.4%. This is very high especially as it was a short time follow-up study. However, since the majority of the babies were clinically normal, term babies with no clinical problem, the parents may have considered follow up unnecessary. The rate of the present study is higher than the 32.6% rate obtained by Wilbur et al8 in a controlled trial study among adult subjects and also than 44.7% by Young et al9 among children. Problems leading to underutilization of health facilities like lack of good, affordable transportation, long distances to health facilities, poor services, high cost of drugs and services and long waiting time have biased the community against the modern health facilities. Other reasons for poor usage include incessant and abrupt stoppage of services through strikes at very critical period of care, ignorance and prevailing poverty.2,3,4,5,6.
Maternal factors found in the present study which favour attrition include lack of antenatal care and mothers from low socioeconomic groups. Primiparous and teenage mothers may also be expected to fail to keep clinic appointments because of their lack of maturity and poor motherhood experience. However, the results are not at a significant level in the present study. These factors are inter-related and could be linked with the prevailing poverty in the developing countries. Hospital services are expensive, transportation is unaffordable to many families especially with the bad roads which make clinic visits time taking. Even if a researcher wishes to help patients to pay or subsidize for the transportation and drugs, he is unlikely to pay for the long time spent on the road to get to the health facilities and for the feeding of the whole family which may be compromised when the mother fails to get to her place of work. Low socioeconomic status may also be related to ignorance or poor understanding of the purposes and extent of a research work. Follow up or cohort research work therefore in the poor nations of the world are prone to be burdensome.
The present study also showed higher proportions of female sex and well babies to be lost to attrition than male sex and ill/admitted babies. Illness in the child, wasting, presence of sequelae may force a parent to look for solutions. Thus babies obviously ill for example, those with fetal malnutrition and other illnesses are likely to be brought up to follow up clinics conversely well babies may not be brought. Gender preference of male sex is a relevant factor in the environment of present study and it may explain the predominance of males in the clinic.6 Attrition may therefore be a particularly important consideration when clinically well individuals are being used as cohorts and control.10
To reduce the problem of attrition in a research study especially when following up apparently normal individuals, education must be thorough and repetitive. A generous allowance should be made for attraction in the estimation of the sample size. Research study should be seen to be relevant to the current needs of the subjects and their socioeconomic conditions.11 The researcher should also consider paying transportation or similar costs of subjects. Efforts should be made by the government to improve current infrastructure including transportation, road network, as well as the socioeconomic status of the families through education and economic empowerment programmes. Involvement of social workers and or health visitors may also be helpful.
Dr. Olusegun Joseph Adebami Department Paediatrics and Child Health, College of Health Sciences, Ladoke Akintola University of Technology, Osogbo, Nigeria. E-mail: firstname.lastname@example.org Phone: +2348037115347.