Population coverage and acceptability of screening programmes

There are a number o f reports o f anaemia screening programmes in pre-school children. All have been based on population blood sampling, ie taking blood samples from all children, and have bad varying degrees o f success. The fundamental problem is ensuring full coverage o f the eligible population. Uptake o f screening offer in reported programmes is summarised in Table 5.1.

Table 5.1 Uptake o f invitation of IDA screening in community based programmes

NB: If final percentage population coverage differs from percentage accepting offer o f screening, due for example to collection o f insufficient blood sample, this is indicated in the comments column.

Author(s) Uptake (%) Comments

James et al (1988) 69 Children aged 1-4 years, n =527. Inner city

practice in Bristol.

James et al (1989) 90 Children 13-24 months, n=136. Routine

screening programme started in response to findings o f above study. Population coverage = 89%.

M ills (1990) 65 Children aged 8-24 months. East London area.

M arder et al (1990) 53 Children aged 15-24 months. Recruited from 2

inner city child health clinics in the

Nottingham area. 58 children recruited from a third clinic, numbers o f eligible children registered with this clinic not given.

Duggan et al (1991) ?23% (see

comments)

Study designed to screen for iron status and dietary intake in Asian children in the Sheffield area. Authors contacted parents o f 310

children out o f an estimated eligible population o f 600 aged 4-40 months. Parents o f 138 children allowed blood samples to be taken.

James et al (1997) Practice 1 100 Practice 2 85 Practice 3 68 Practice 4 93 Practice 5 100 All practices 89

Study to assess feasibility o f rolling

programme out to other practices, n=136,177, 84, 97 & 45 respectively. One practice (3) also had particular problems with collection o f blood samples. Estimated coverage o f

Author(s) Uptake (%) Comments

for practices 1-5.

Publication also refers to an unsuccessful and unpublished attempt by one o f the authors to instigate screening via child health clinics in an inner city area o f London. Uptake was less than 10% (n=600).

Moy & Aukett (1998)

56 Authors do not present data as a percentage o f

all children eligible for screening ie all those eligible for 21 month development check, but as a percentage o f all those who attended a clinic for the development check (n=625). Birmingham area.

Tamhne (1998) 58 Invitations sent to 90 families, 52 families

volunteered. Blood samples taken from only 16 o f the 23 Asian children and 6 o f the 29 non-Asian children. Leicester area.

Goodhart & Logan (1999)

81 Children aged 12-24 months, n=335. North­

east London. Problems with blood collection, therefore final population coverage = 51%.

Uptake in these studies was highly variable ranging from 100% (James et al 1997) to such low levels that the programme was suspended (Tamhne 1998). There are no established criteria for a judging a programme to be successful in terms o f uptake although logically it should be in excess o f 90%. M oy & Aukett (1998) described the screening programme in Birmingham as “operating successfully”, however, more than a third o f eligible children were not screened. James et al (1997) reported large variations in population coverage over five practices. Reasons for not taking up offer o f screening included; carer was not the parent so felt they could not give permission (Moy & Aukett 1998), not wanting child to have a painful/traumatic procedure or parental assertion that the child was unlikely to be anaemic (Mills 1990, M arder et al 1990, James et al 1997) and language difficulties (Marder et al 1990). James et al 1997 suggested that failure to achieve full population coverage together with other problems make it unlikely that population screening in inner city type areas would be a successful strategy for detecting IDA.

The figures in Table 5.1 refer to uptake o f offer o f screening. In some cases the final numbers screened were further reduced by inability to test samples. For example, Marder et al (1990) reported that 13% o f collected samples were either insufficient in

volume or clotted. Moy & Aukett (1998) reported that failures to test blood samples were largely due to unavailability or malfunctioning o f the baemoglobinometer.

bi addition to the problems outlined above, there is an uncertainty about what age(s) children should be screened. Ideally screening would be undertaken and IDA treated before potential problems (see Chapter 3) develop. Wharton (1999) has pointed out that if screening is at the age o f greatest prevalence ie about 18 months, some children will have been anaemic for months and others will go on to develop anaemia at a later time. There is little information about the natural history o f IDA although work by James et al (1993, 1995) indicated that iron status at 14 months was not predictive o f iron status at 2 years. The authors concluded that screening at 14 months was an ineffective method o f identifying those at risk o f IDA during early childhood. Conversely a longitudinal study o f Irish children found that haemoglobin concentration at 12 months was a significant predictor o f haemoglobin at 24

(p=0.001) and 36 months (p=0.002) (Freeman et al 1998). A recent study that

measured iron status and developmental outcome longitudinally from birth onwards in a large cohort o f children, suggested that the most appropriate time for screening would be at 8 months old or earlier (Sherriff et al 2001). This was based on the finding that lower haemoglobin concentration at 8 months was associated with poorer performance in motor development tests at 18 months old.

5.1.2 Comment

It is clear there are a variety o f problems with the practicalities o f running IDA screening programmes. Even the authors o f the longest running programme have concluded that population screening (ie taking blood samples from all children) is an unsuccessful strategy for detecting anaemia (James et al 1997). It is also important to note that screening programmes use considerable resources within general practices. Robson (1998) lists IDA as just one o f 41 conditions that could be screened for in primary care. Despite the above, IDA meets the WHO criteria as a condition that is

suitable for screening. Peckham & Dezateux (1998) summarised these criteria

complete with later revisions to cover “contemporary screening issues”. According to those criteria IDA should pose a health problem, its natural history should be understood and it should be detectable at a latent or early symptomatic stage. It is the position o f this thesis that the balance o f evidence suggests that IDA is associated

with motor developmental delay, probably with mental development and that the relationship is probably causal. IDA therefore represents a health problem for young children (see Chapter 3). Its natural history is somewhat unclear but there is recent information (Sherriff et al 2001) about longitudinal iron status in a large group o f young children. It is difficult to assess the true natural history o f the condition simply because any study designed to measure iron status would be ethically bound to offer treatment to anaemic children therefore altering the natural course o f the condition. Finally, blood testing will detect anaemia at an asymptomatic stage. All studies o f IDA prevalence have been in children who are ostensibly healthy and yet it is detected at rate o f 12-39% (see section 2.5). By these criteria IDA is a suitable condition to screen for but for any future programme to be successful it must be based on a method other than blood testing to determine which children are likely to be anaemic.