The present study provides information on neurodevelopmental outcome in a cohort of VLBW infants in Johannesburg, South Africa. The mean BSID 111 scores at a median corrected age of 16.48 months age were within normal limits for the cognitive, motor and language subscales. Furthermore, considering 70 as the cutoff for the BSID 111 score, only 15.1% of infants had an abnormality on a single subscale, while 3.7% were affected in all 3 areas. However, the mean score on each subscale was less than the anticipated population mean of 100. Almost one third of patients were considered to be "at risk", with a BSID 111 score between 70 and 85, indicating that this is a high risk group of children likely to have long term developmental problems who warrant ongoing monitoring and intervention. Currently long term follow up of ex preterm infants in developing countries is frequently limited to those with obvious handicap due to insufficient resources.
This group of VLBW infants represents a select group, as ventilatory support was not offered to infants with a birth weight less than 900 grams, resulting in a very low survival of infants of lower birth weight. Although other neonatal units in developing countries may not apply strict birth weight cut offs for ventilation, or the birth weight cut offs may be different, the challenges of being unable to ventilate all preterm infants who require support will be very similar.
The neurodevelopmental outcome results in the present study are similar to those of Cooper and Sandler [15
] at Chris Hani Baragwanath hospital, Soweto, in the 1990s. They found 15.3% of VLBW infants had an abnormal BSID score at follow up. The results are also comparable with those reported in the literature. In a follow up study reported from Bangladesh [14
], only 32% of VLBW infants were reported as neurodevelopmentally intact at 12 months of age. Many follow up studies are confined to extremely premature infants, born between 22 and 26 weeks gestation. Up to one quarter of these infants will have at least one major disability in childhood [21
]. The rates of disability in VLBW infants or those born after 26 weeks gestation are heterogeneous. Of a group of babies born < 29 weeks gestation between 1985 and 1987, only 31% had no physical or educational handicap and 21% had at least one severe disability at 7 years of age [9
]. A group of infants born at a median age of 28 weeks and assessed at a median corrected age of 18 months showed normal outcome in 59%, borderline function in 26% and abnormal outcome in 15% [22
] The survival without neurodevelopmental disability of infants born < 30 weeks gestation improved from 62% in infants born between 1985 and 1986 to 81% for those born between 2005 and 2006 [23
]. In the EPIPAGE study [3
] in France, infants born below 32 weeks gestation between 1997 and 2001 showed normal profiles at 6 to 10 years of age in 68%, minor disorders in 18% and major disorders in 14%. A group of Finnish infants born between 2001 and 2006 were assessed at 2 years of age and 9.9% were found to have neurodevelopmental impairment. The outcome of preterm infants below 32 weeks born at a tertiary centre in Ankara in Turkey, 16.6% were found to have minor neurological dysfunction and 8.3% to have cerebral palsy at a median age of 25.85 months [24
]. Of this group, 24.8% had a low Bayley Psychomotor development index and 25.4% a low Bayley Mental Development index.
In the present study, cystic PVL was associated with poor cognitive, motor and language function. Further, duration of supplemental oxygen, prolonged hospitalisation, resuscitation at birth and increased maternal parity were associated with poor outcome. Duration of intensive care showed a trend towards worse outcome. These findings were similar to those reported in other studies where NEC [24
], gender [7
], chronic lung disease [3
], respiratory distress [35
], multiple birth [7
], HIV infection [37
], cranial sonar findings [29
], particularly PVL [15
] and intraventricular haemorrhage, [26
]), neonatal seizures [26
], perinatal asphyxia [41
], neonatal sepsis [27
], postnatal steroids [31
] and the duration of assisted ventilation [24
] have all been associated with adverse neurodevelopmental outcome in VLBW infants. Gestational age [3
] and birth weight [10
] have both been reported as predictors of poor neurodevelopmental outcome. This was not the case in the present study, which almost certainly reflects the ventilation policy in the unit at the time, where babies with a birth weight below 900 grams were not offered ventilation, so survival in this birthweight category is low [47
]. Other neonatal factors such as treated hypotension [48
] were not associated with poor outcome in the current study - possibly due to small numbers of affected patients. HIV exposure, ethnicity and KMC were not predictive of outcome in the present study. The present study shows no difference between males and females with regard to neurodevelopmental outcome, which is contrary to findings in other studies where male gender is associated with worse outcome [29
]. The reason for this is unclear and would have to be confirmed in future research.
Limitations of the study
The rate of cerebral palsy (CP) is of great importance when considering the outcome of preterm infants. The reported number of CP cases in the present study is low (3.7%). This could be an underestimation of the true rate of cerebral palsy in this population as the age of assessment in the present study is too low to reliably report on the rate of cerebral palsy. The diagnosis of cerebral palsy in many children can only be made reliably after the age of 2 years [49
]. A significant number of preterm infants followed up in one study had a change in neurological diagnosis made at 18 months as compared to 30 months and it therefore may be necessary to delay the diagnosis of cerebral palsy in some children [50
]. It is likely that those infants, who present later, will be relatively mild in comparison to those who present early. It is also possible that some cases of CP were among those lost to follow up. However, the CP rate in this population would be expected to be low, as the sickest and smallest of these infants do not survive [47
Although the rate of follow up achieved of 74.6% is acceptable and comparable to other reports, it is possible that a number of handicapped children were lost this way. The most common reason for non compliance was relocation of the parents to their place of origin, some as far afield as Malawi and Tanzania. Parents also return to work and find it difficult to bring their children for follow up after the first few visits. Transport and hospital strikes also resulted in the loss to follow up of some patients. Five patients returned to follow up on the incorrect day when a physiotherapist was unavailable, so did not have a Bayley assessment; these children were all clinically normal.
The BSID 111 has been tested on black South African children, between 0 and 18 months of age, who did not have risk factors or pre-existing conditions. The results showed that these children performed well and were often above average (a composite score > 100 on each BSID 111 subscale), confirming that the BSID 111 is suitable for use in line with previous studies in this population [51
Another limitation of the study is that children had assessments done at different ages for the reasons outlined above. Ideally all assessments should have been done at the same age on all patients. As previously noted, the age of assessment influences the result obtained in the BSID scores. For this reason and high missing data on follow-up visits, we reported the latest BSID 111 in each patient as the outcome variable and did not conduct a repeated measures analysis, which would have been ideal.