Can Western developmental screening tools be modified for use in a rural Malawian setting?

Transcription

1 Additional figures and a table are published online only at vol93/issue1 1 Department of Paediatris, College of Mediine, Blantyre, Malawi; 2 Postgraduate Statistis Centre, Department of Mathematis and Statistis, Lanaster University, Lanaster, UK; 3 Centre for Medial Statistis and Health Evaluation, University of Liverpool, Liverpool, UK; 4 Department of Community Health, College of Mediine, Blantyre, Malawi; 5 Department of International Health, University of Tampere Medial Shool, Finland and Department of Paediatris, Tampere University Hospital, Finland; 6 Institute of Child and Reprodutive Health, University of Liverpool, Liverpool, UK Correspondene to: Dr Melissa Gladstone, Institute of Child Health, University of Liverpool, Royal Liverpool Children s Hospital, Eaton Rd, Liverpool L12 2AP, UK; mgladstone@btinternet.om Aepted 4 Marh 2007 Published Online First 22 Marh 2007 Can Western developmental sreening tools be modified for use in a rural Malawian setting? M J Gladstone, 1 G A Lanaster, 2 A P Jones, 3 K Maleta, 4 E Mtitimila, 5 P Ashorn, 5 R L Smyth 6 ABSTRACT Objetive: To reate a more ulturally relevant developmental assessment tool for use in hildren in rural Afria. Design: Through fous groups, piloting work and validation, a more ulturally appropriate developmental tool, based on the style of the Denver II, was reated. Age standardised norms were estimated using 1130 normal hildren aged 0 6 years from a rural setting in Malawi. The performane of eah item in the tool was examined through goodness of fit on logisti regression, reliability and interpretability at a onsensus meeting. The instrument was revised with removal of items performing poorly. Results: An assessment tool with 138 items was reated. Fae, ontent and respondent validity was demonstrated. At the onsensus meeting, 97% (33/34) of gross motor items were retained in omparison to 51% (18/35) of soial items, and 86% (69/80) of items from the Denver II or Denver Developmental Sreening Test (DDST) were retained in omparison to 69% (32/46) of the newly reated items, many of these having poor reliability and goodness of fit. Gender had an effet on 23% (8/35) of the soial items, whih were removed. Items not attained by 6 years ame entirely from the Denver II fine motor setion (4/34). Overall, 110 of the 138 items (80%) were retained in the revised instrument with some items needing further modifiation. Conlusions: When reating developmental tools for a rural Afrian setting, many items from Western tools an be adapted. The gross motor domain is more ulturally adaptable, whereas soial development is diffiult to adapt and is ulturally speifi. Eighty perent of the world s disabled population live in low inome ountries, many of these in Afria. 1 The World Health Organization has made early identifiation of hildren with disabilities a high priority, espeially as early rehabilitation may redue the impat of impairments. 2 3 To identify these hildren and provide basi servies, developmental milestones need to be learly identified. Furthermore, linial studies investigating interventions in hildren require normal parameters. When hild development is assessed in linial studies in developing ountries, Western developmental tools are often utilised. 4 5 These inlude the Bayley sales, 6 the Griffith s, 7 the MCarthy sales 8 and the Denver II, 9 all designed and validated in Western ountries. These tools may be tailored for use in non-western settings. Often translation (hanging of the language used) is all that is arried out If this is not aompanied by a proess of adaptation, translation alone may not allow ompletely for loal expressions and ustoms, Original artile therefore leading to misinterpretation of results. 12 In other settings, tools are adapted and items are modified and in some ases new items are reated for use within a Western tool. 13 Sometimes these tools are piloted (tried out before use) 14 and validated (assessed that they are measuring what they are supposed to be measuring) in the loal population. 15 Even these adapted tools, however, are of limited value without normal ranges for their defined population. Standardisation studies (finding norms for a population) have taken plae in many non-western ountries mainly using the Denver Developmental Sreening Test (DDST) in a translated and oasionally adapted form, but none of these studies was in Afria. Only two studies have attempted standardisation in Afria, one using a translated form of the Bayley sales with an urban blak South Afrian population 19 and the other on a limited age range in a rural Nigerian population. 20 It is lear that Western developmental assessment tools may inlude tasks and materials whih are ompletely alien to other ultures. These tools may therefore fail to identify and assess hildren adequately in ultural settings other than those for whih they were reated. 21 This may be less of a problem when omparing groups of hildren, but when Western tools are used alone as an outome measure, ulture may have an effet. In theoretial studies, ulture has been demonstrated to have an influene on hild development, partiularly in the area of soial development Cognitive abilities suh as memory, ategorisation tehniques and pattern reognition have also been reported to be influened by ulture Even gross motor development may possibly be affeted by ulture In this study, we aimed to reate a simple, ulturally appropriate developmental assessment tool adapted and modified from Western tools and standardised for use in rural Malawi. The first stage in the development of this tool was to identify whih items from Western tools (eg, the DDST or Denver II) were not relevant to the ageappropriate experienes of rural Malawian hildren. These items were then replaed with ones more appropriate to this ultural ontext. We did this firstly by holding fous groups to agree whih items should be replaed and to reate alternative items. All items (both retained and new) were then validated and standardised in a large population study. The performane of all items was examined in a onsensus meeting and a revised instrument proposed. Arh Dis Child 2008;93: doi: /ad

2 Figure 1 The proess of reating a more ulturally appropriate developmental assessment tool. METHODS Setting and study population This study was a substudy of the Lungwena Child Survival Study (LCSS), a prospetive family ohort study looking at gestational health and the growth, development, morbidity and mortality of rural Malawian infants and hildren. Lungwena is an area in southern Malawi where a government health entre serves an approximately 100 km 2 rural area with some people in 23 villages. Most of the inhabitants are Muslims of the Yao tribe. The literay rate is low and subsistene farming and fishing are the main oupations. The original ohort for the LCSS was enrolled between June 1995 and August All pregnant women presenting for antenatal are were eligible for the LCSS and 97% of the population of pregnant women in the area, at that time, were enrolled in the study. Details of reruitment, olletion of bakground data and follow-up have been desribed previously. The population of hildren used for this study is the original LCSS ohort of hildren aged years and younger siblings aged years. Out of the 1237 LCSS hildren and siblings available, 1197 were seen, with 40 families either refusing to take part or not being available. The ages of the hildren were known from LCSS birth data or from the health passport given to mothers at the birth of their baby where the date of birth is reorded and whih almost all mothers arry with them for all health appointments. A quota sampling strategy was used as in the DDST and Denver II 34 with target numbers of hildren being sought in eah of 33 age groups (see supplementary table A). A total of 67 hildren were exluded due to premature birth (34 weeks or less measured by fundal height at the antenatal lini), 32 twin birth or signifiant disability inluding severe malnutrition (weight for height z sore of less than 22), leaving 1130 hildren in the final analysis. The LCSS reeived approval from the National Health Siene Researh Committee in Malawi (HSRC 93/94). Informed verbal onsent was sought from eah mother at the beginning of the LCSS and again before a development assessment was arried out. Creation of the developmental assessment tool The Denver II, DDST and Griffith s instruments were examined by the Malawian researh team. Items onsidered to be ulturally appropriate were inluded and translated, whereas those onsidered inappropriate (suh as prepares ereal or plays board/ ard games ) were removed. New items and modifiations to 24 Arh Dis Child 2008;93: doi: /ad

3 Table 1 Domain of development Examples of speifi items added or removed during the proess of reating new more ulturally appropriate tool DDST or Denver II items removed at stage 1 New items replaed at stage 1 Western test items were then reated through disussions with a series of fous groups. Key informants were the eight loal researh workers. They were all women of hild-bearing age with at least 8 year s eduation and researh experiene of at least 5 years. Themes relating to developmental milestones were disussed and ideas from these sessions were used to reate new items. Illustrations were made for most items in the instrument and used as prompts for the researh workers. Some ame with permission from Disabled village hildren. 35 Fae validity 36 and ontent validity were assessed by all researh assistants, five Malawian paediatriians, a language expert from the University of Malawi and six medial students at the College of Mediine, Malawi. One the new instrument was reated, the team was trained in its use and it was piloted in two stages. At eah stage, feedbak and training were given and problemati items were re-adapted or re-translated. The proess of reating and refining the more ulturally appropriate tool is shown in fig 1. Standardisation using a normal population sample Overall, 1197 hildren were assessed on one oasion between February 2000 and April 2001 on a home visit by researh Denver II/Griffith s items removed or needing modifiation after onsensus meeting New items removed or needing modifiation after onsensus meeting Gross motor 1. Head up to 45u 1. Danes to musi 1. Pulls self to stand (not 1. Danes (not speifi enough) 2. Walks up steps 2. Kneels well to be respetful understood by researh 2. Carries objet on head with one hand 3. Pedals triyle 3. Climbs and gets down assistants) (need to speify exat objet) 4. Carries objet on head with 2. Touhes toes with legs 3. Carries water on head with no hands one hand straight (diffiult to teah) (need to speify size) 5. Pushes wheel with stik 6. Carries water on head no hands Fine motor 1. Looks for yarn 1. Bangs drum/pot 1. Draws a person with 3 parts 1. Moulds ball with lay (task not 2. Thumb wiggle 2. Moulds ball with lay (many not passing by age 6) normally done by hildren) 3. Thumb finger grasp 3. Makes doll of lay 2. Draws a person with 6 parts 2. Bangs drum/pot (translation plays 4. Sorts objets into groups (many not passing by age 6) drum inorret) 5. Piks longest stik 3. Builds house of orn obs (sex speifi) Language 1. Responds to bell 1. Sings 1. Knows 3 olours (olours 1. Uses respetful terms (poor reliability) 2. Squeals 2. Knows today and tomorrow used differently in Malawi) 2. Knows father s name (poor reliability) 3. Points to pitures/name 3. Knows name of own village 2. Knows opposites (not speifi 3. Can ount bottle tops to 10 (ounting pitures 4. Uses respetful terms enough question needed learned at shool not many hildren 4. Counts 1 blok 5. Knows father s name larifying) in shool/nursery aged under 5) 5. Uses plurals 6. Can ount bottle tops to 10 Soial 1. Plays pat-a-ake 1. Shares things 1. Helps in house 1. Spends most of time on mum s bak 2. Plays ball with examiner 2. Plays with friends 2. Puts on lothing with help (family dependent) 3. Resists toy pull 3. Spends most of time on (poor reliability) 2. Shy with strangers (poor reliability) 4. Plays peek-a-boo mum s bak 3. Feeds self (speify what 3. Sweeps (sex speifi) 6. Uses spoon/fork 4. Eats with family in a group type of food) 4. Washes dishes (sex speifi) 7. Feeds doll 5. Can plant seeds by self 4. Drinks from up by self 5. Goes to the toilet with help (not speifi 8. Brushes teeth with help 6. Washes dishes (not speifi enough enough, eg, older hildren need help using 9. Names a friend 7. Washes lothes regarding how well or not pit latrines) 10. Plays board games 8. Sweeps hild needed to do this) 12. Prepares ereal 9. Goes to the toilet with help 13. Plays interative games, eg, tag 10. Goes to the toilet by self 14. Buttons up 11. Makes porridge by self 12. Adds wood to fire 13. Can build fire by self assistants. The assessment took approximately 35 min to omplete and where possible, items were diretly observed. In a few ases a report was given, for example does he go to the toilet by himself?. Items were sored as either pass or fail, or don t know if the hild was unooperative or unwell. Items were asked until the hild failed seven items in a row. Data entry and analysis were arried out using Mirosoft Exel 6.0, SPSS 11.1, Stats-diret and STATA omputer programs. Eah hild in the study was identified by a ode. Data were heked prior to analysis and any outlying results were reviewed. Standardisation is the proess of determining normal age ranges for whih hildren pass the items for a developmental assessment tool. A logisti regression analysis was arried out with deimal age and sex as explanatory variables. The observed and predited probability of passing was determined and graphs were drawn for eah item. The goodness of fit of the graph was visually assessed and disrepanies reviewed. To determine statistially whether or not the fitted urve was a suffiiently good representation of the data, a goodness of fit statisti was alulated. 37 If this was signifiant at the 5% level, indiating a poor fit, then the data were re-examined and refitting was done Arh Dis Child 2008;93: doi: /ad

4 Figure 2 Examples of (A) a good logisti fit for fine motor question 8 Transfers objets from hand to hand, (B) spline fit for gross motor question 17 Walks bakwards, and (C) a poorly worded question (soial question (SOC) 16 Can put lothes on with help ). using triple split spline regression. The ages orresponding to the 35th and 65th perentiles were alulated from the original fit to determine the ut-points. For some items that performed less well, the ut-points were hosen by viewing the graphs to failitate a good fit. Three logisti urves were then fitted, one for eah region, based on the split Any items with signifiant gender effets were removed or onsidered for further modifiation to ensure the tool was appliable to all hildren irrespetive of gender. Using the predited probabilities found from the logisti regression analyses, the ages orresponding to 25%, 50%, 75% and 90% of the hildren passing were determined for eah item. These were then used to plot the age norms of ahievement of eah milestone in a box-type representation. Reliability of the items Reliability for eah item was tested by using two subsamples of 60 (inter-observer) and 28 (intra-observer) randomly seleted hildren who were seen at 7 and 14 days after initial assessment. Of the 60 hildren, 46 ompleted the follow-up using two different examiners (inter-observer), while 25 of the 28 hildren used the same examiner (intra-observer). All items in the tool were assessed for both types of reliability. Kappa statistis (k) with 95% onfidene intervals (CI) were used to alulate the degree of observer agreement for eah question. Positive values of 0 to,0.2 indiate poor agreement,.0.2 to 0.4 fair agreement,.0.4 to 0.6 moderate agreement,.0.6 to 0.8 good agreement and.0.8 to 1 very good agreement. 40 Respondent validation was arried out after the preliminary analysis. This method of validation involves the reporting of findings bak to the partiipants. Findings were fed bak at the end of the study to the Lungwena Health Centre Management Committee. This onsisted of four hiefs, one overall representative and three women representatives, all from the loal area. Consensus meeting One all the items were analysed, an expert panel (MG, AJ, EM and GL), whih inluded a Malawian paediatriian, met to review the results and deide whih items should remain, whih should be modified and whih should be removed. Items were judged on their graphial representation, and goodness of fit on logisti regression, reliability and subjetive ratings of interpretability by partiipants and researhers. RESULTS A tool with 138 items (34 gross motor (GM), 34 fine motor (FM), 35 language and 35 soial items) was reated. An example of the tool is shown in supplemental fig B (see supplementary data). Most (58%) items were from the DDST and Denver II, with a small perentage (9%) from the Griffith s instrument. Many items in the GM (82%), language (77%) and FM (70%) setions were diretly translated from Western tests with modifiations mainly in the FM setion. Only 37% of the soial items were taken from Western tests. The first two olumns of table 1 provide examples of many of the items that were removed from the DDST or Denver II, and shows the newly reated items whih replaed them. The fae validity and ontent validity of the tool were tested. The modified instrument appeared to those questioned to over development in hildren in ways that were important, and it was judged to examine in a fully omprehensive and logial fashion the domains of hild development for hildren in Malawi. It was therefore onsidered to have good fae validity and ontent validity. Most items were found to be aeptable for studying hildren s development in this setting through respondent validity. The pitures as prompts were found to be partiularly helpful to the researhers in the field. Examples of graphs reated through logisti regression during the standardisation proedure and where triple split joined 26 Arh Dis Child 2008;93: doi: /ad

5 Figure 3 Example of developmental milestones ahieved by Lungwena hildren in the area of language development. Age ranges are given for perentage of hildren passing an item. regression was used, are shown in fig 2. In terms of goodness of fit on logisti regression and on spline regression, soial items had the highest number of poor fits (51%, 18/35), sex being an independent preditor in some of these (23%, 8/35) (see table 2). A larger proportion of the newly reated items had a poor fit on logisti regression (15%, 7/46) and an effet of sex (17%, 8/46) than those from the Western tools. The few items not attained by 6 years ame from the fine motor area of development and inluded draws a man with 6 parts and draws a square. The results of the Lungwena milestones for the language setion of development are shown in fig 3. The other areas of development are desribed in supplemental fig C (see supplementary data). Table 2 Reliability results are also shown in table 2. For inter-observer reliability, 82% (113/138) of the questions had moderate to very good reliability (k.0.4). There are no figures in the Denver tehnial manual for inter-observer reliability for omparison. Intra-observer reliability demonstrated moderate to very good reliability (k.0.4) for 75% (106/138) of the questions. This ompares well with Denver II figures, 34 where 81% of their items had a k.0.4. In relation to the domains of development, GM items had the best overall inter-observer (29/34 items) and intra-observer (32/34 items) reliability with k.0.4. Items from the soial area performed less well, with only 74% (26/35) of the items on inter-observer Deisions regarding suitability of items within eah domain of development and the soure of the question Domain of development Soure of question Denver GM FM Lang Soial or DDST Griffith s Newly derived No. of items in total Inter-observer items with k Intra-observer items with k Poor fit on logisti regression Sex effet on logisti regression Not attained by 6 years Poor interpretability or translation To be inluded or modified for revised version of the tool DDST, Denver Developmental Sreening Test; FM, fine motor; GM, gross motor; Lang, language. Total Arh Dis Child 2008;93: doi: /ad

6 reliability and 60% (21/35) of the intra-observer items having a k.0.4. In relation to the soure of the item, more of the loallyderived items had poor inter-observer (33/46) and intra-observer (15/46) reliability (k,0.4) in omparison to those items derived from the Denver II (12/80 and 8/80). After a onsensus meeting, 110 of the 138 items (80%) were retained in the revised instrument, with some needing further modifiation. Only 69% (32/46) of the newly reated items were retained in omparison to 86% (69/80) of the DDST or Denver items used (see table 2). The results of this meeting giving examples of items removed are detailed in the last two olumns of table 1. DISCUSSION We have demonstrated that many items from Western tools an work well when adapted and translated for other settings. They have already had their own rigorous reliability and validity studies arried out in the West and therefore are more likely to be robust in use. However, through our fous group, validation and piloting work, we have also demonstrated that in all domains of Western tests (suh as the DDST), there are some items whih are ulturally inappropriate for a rural Afrian population. For example, questions suh as prepares ereal or plays board games/ard games are unommon ativities for hildren in rural Afria. Also, the pink doll in the DDST kit was terrifying to most hildren when used in piloting; many hildren had never seen anything like it and many sreamed. It would have been unlikely that we would have been able to get them to sit down and feed the doll. Some of the naming questions in the Language setion of the DDST or Denver II have pitures of objets that hildren, at least in the part of rural Afria studied, have never seen before, suh as a horse and a ar. This makes it diffiult for them to name them, espeially as many hildren have also never seen a book at their age, or pitorial representations of many objets. In the reation of new items, however, many newly reated items were less reliable, more sex-speifi and had poorer goodness of fit in logisti regression. This was most evident in the soial domain and least evident for gross motor skills. Soial skills seem to have the least universality and in measuring them, we need to question the appropriateness of the onepts being measured in suh different settings. When measuring soial skills we may be determining the ability of the hild to have learned important skills instilled by parents and arers in partiular ultural settings, but this an only be measured if pertinent skills are tested for. The diffiulty when reating new soial items for a tool suh as this, is that the items must be speifi enough to distinguish between the developmental age ranges of hildren, but also be lear and easy to explain in a developmental tool. This will ontinue to be a hallenge. It was not a primary aim of our study to ompare our results with the Denver II or DDST. A formal statistial omparison has not been possible; however, when omparing our harts with those from the Denver II or DDST on gross omparison, it does seem that there are obvious differenes in milestones with hildren from the West. For example, the item ombines two words in the Denver II is attained at between 17 and 21 months, whereas in our sample this was obtained at between 21 months and 2 years 4 months. This demonstrates the importane and neessity of reating norms for a given Afrian population, as they are likely to be different from those in the West. A seond phase of work is urrently underway using the methodology that we have formulated in this first study to What is already known on this subjet The ages of attainment of developmental milestones an differ with the ultural bakground of the hild, although there are very few data from Afria. Developmental assessment in Afrian settings is often arried out using translated and adapted Western tools, with a dearth of tools adapted, validated and standardised speifially for rural Afrian hildren. What this study adds A proedure using validity, reliability, goodness of fit on logisti regression and expert onsensus has been devised for assessing newly reated questions or items adapted or taken from Western developmental assessment tools. Many items from Western developmental assessment tools an work in a rural Afrian setting. Gross motor items are most reliable, whereas items within the soial area of development need to aount for ultural differenes. refine a further tool with a larger standardisation sample. This work will inlude reating a soring system, and arrying out more detailed reliability measurements and further validity tests of between-group and onstrut validity. One this new version has been reated and has undergone the strit proedures that we have instituted in our methodology, we hope to have reated a tool that may benefit ommunity health workers in other rural settings in Afria after loal validation. The omplete tool may also be used by researh workers who are investigating developmental outomes as part of their intervention strategies. Aknowledgements: Our thanks to all the staff and researh assistants at the Lungwena Child Survival Study site in Lungwena, Malawi, to Professor R Broadhead, head of the Department of Paediatris at the College of Mediine in Blantyre, Malawi, to Professor MB Duggan, professor in paediatris and ommunity health who worked at the College of Mediine in Malawi when the study was arried out for her help and suggestions, and Dr L Rosenbloom, reently retired onsultant paediatri neurologist at the Royal Liverpool Children s Hospital, for his help and suggestions. Our thanks also to Dr W Frankenberg and H Shapiro for omments as well as statistial advie when writing up the study. Funding: Funding was supplied by the Aademy of Finland, the Foundation for Paediatri Researh in Finland, the Medial Researh Fund of Tampere University Hospital, the Alexander Wernher Piggot Memorial Trust and the EWG Memorial Trust. There was no involvement from the study sponsors in the study design, olletion, analysis, interpretation, writing of reports or deisions to submit. Competing interests: None. REFERENCES 1. WHO. Disability, inluding prevention, management and rehabilitation. Report by the Seretariat, (aessed 21 Otober 2007). 2. Committee on Children with Disabilities. Developmental surveillane and sreening of infants and young hildren. Pediatris 2001;108: Durkin MS, Davidson LL, Desai P, et al. Validity of the ten-question sreen for hildhood disability: results from population based studies in Bangladesh, Jamaia and Pakistan. Epidemiology 2005;5: Powell CA, Walker SP, Himes JH, et al. Relationships between physial growth, mental development and nutritional supplementation in stunted hildren: the Jamaia study. Ata Paediatr 1995;84: Joos SK, Pollitt E, Mueller WH, et al. The Baon Chow Study: maternal nutritional supplementation and infant behavioural development. Child Dev 1983;54: Bayley N. Bayley Sales of Infant Development manual. 2nd ed. San Antonia: Harourt Brae, Griffiths R. The abilities of young hildren. Amersham: ARICD, Arh Dis Child 2008;93: doi: /ad

7 8. MCarthy D. MCarthy Sales of Children s Abilities. New York: Psyhologial Corporation, Frankenburg WK, Dodds J, Arher P, et al. The Denver II: a major revision and restandardization of the Denver Developmental Sreening Test. Pediatris 1992;89(1): Boivin MJ, Green SD, Davies AG, et al. A preliminary evaluation of the ognitive and motor effets of pediatri HIV infetion in Zairian hildren. Health Psyhol 1995;14(1): Boivin MJ, Giordani B, Ndanga K, et al. Effets of treatment for intestinal parasites and malaria on the ognitive abilities of shoolhildren in Zaire, Afria. Health Psyhol 1993;12(3): Geisinger K. Cross-ultural normative assessment: translation and adaptation issues influening the normative interpretation of assessment instruments. Psyhol Assess 2006;6(4): Drotar D, Olness K, Wiznitzer M, et al. Neurodevelopmental outomes of Ugandan infants with HIV infetion. An appliation of growth urve analysis. Health Psyhol 1999;18(2): Stoltzfus RJ, Kvalsvig JD, Chawaya HM, et al. Effets of iron supplementation and antihelminthi treatment on motor and language development of preshool hildren in Zanzibar: double blind, plaebo ontrolled study. BMJ 2001;323: Holding PA, Taylor HG, Kazungu SD, et al. Assessing ognitive outomes in a rural Afrian population: development of a neuropsyhologial battery in Kilifi distrit, Kenya. J Int Neuropsyhol So 2004;10(2): Al-Naquib N, Frankenburg WK, Mirza H, et al. The standardization of the Denver Developmental Sreening Test on Arab hildren from the Middle East and north Afria. J Med Liban 1999;47(2): Lim HC, Chan T, Yoong T. Standardisation and adaptation of the Denver Developmental Sreening Test (DDST) and Denver II for use in Singapore hildren. Singapore Med J 1994;35(2): Song J, Zhu YM, Gu XY. Restandardization of Denver Developmental Sreening Test for Shanghai hildren. Chin Med J (Engl) 1982;95(5): Rihter LM, Griesel R, Rose C. The Bayley Sales of Infant Development: a South Afrian standardization. S Afr J Oup Ther 1992;22(1): Ogunnaike OA, Houser RFJ. Yoruba toddlers engagement in errands and ognitive performane on the Yoruba Mental Subsale. Int J Behav Dev 2002;26(2): Olade RA. Evaluation of the Denver Developmental Sreening Test as applied to Afrian hildren. Nurs Res 1984;33(4): Aess a vast information database with Toll-Free linking 22. Levine RA, Dixon S, Levine S, et al. Child are and ulture: lessons from Afria. Cambridge: Cambridge University Press, Whiting B, Whiting WJ. Children of six ultures. A psyho-ultural analysis. Harvard, MA: Harvard University Press, Kambalametore S, Hartley S, Lansdown R. An exploration of the Malawian perspetive on hildren s everyday skills: impliations for assessment. Disabil Rehabil 2000;22(17): Serpell R. How speifi are pereptual skills? A ross-ultural study of pattern reprodution. Br J Psyhol 1979;70: Rogoff B, Waddell KJ. Memory for information organized in a sene by hildren from two ultures. Child Dev 1982;53: Curran HV. Nigerian hildren: developmental perspetives. London: Routledge & Kegan Paul, Glik J. Cognitive development in ross-ultural perspetive. In: F Horowitz, ed. Review of hild development researh. Chiago: University of Chiago Press, 1975: Super CM. Environmental effets on motor development: the ase of Afrian infant preoity. Dev Med Child Neurol 1976;18: Warren N. Afrian infant preoity. Psyhol Bull 1972;78: Geber M. Developpement psyho-moteur de l enfant afriain. Courier 1956;6: Vaahtera M, Kulmala T, Ndeka M, et al. Ante and perinatal preditors of infant mortality in rural Malawi. Arh Dis Child Fetal Neonatal Ed 2000;82:F Maleta K, Virtanen S, Espo M, et al. Timing of growth faltering in rural Malawi. Arh Dis Child 2003;88: Frankenburg WK, Dodds JB, Arher P, et al. Denver II tehnial manual. Denver, CO: Denver Developmental Materials, Werner D. Disabled village hildren - a guide for ommunity health workers, rehabilitation workers, and families. Berkeley, CA: The Hesperian Foundation, Eiser C, Morse R. Quality of life measures in hroni diseases of hildhood. Health Tehnol Assess 2001;5(4). 37. Lemeshow S, Hosmer D. A review of goodness of fit statistis for use in the development of logisti regression models. Am J Epidemiol 1982;115(1): Greenland S. Dose-response and trend analysis in epidemiology: alternatives to ategorial analysis. Epidemiology 1995;6(4): Pastor R, Guallar E. Use of two-segmented logisti regression to estimate hangepoints in epidemiologial studies. Am J Epidemiol 1998;148(7): Landis R, Koh G. The measurement of observer agreement for ategorial data. Biometris 1977;33: Toll-free linking gives you immediate aess to the full text of many of the ited artiles in a paper s referene list FOR FREE. With the support of HighWire s vast journal atalogue, a huge referene library is now open to you. If HighWire hosts the journal, you an view the full text of the referened artile, ompletely free of harge by following the Free Full Text links in the referenes. Arh Dis Child 2008;93: doi: /ad