Author's response to reviews Title: Validity and reliability of a structured interview for early detection and risk assessment of parenting and developmental problems in young children: a cross-sectional study Authors: Henk F van Stel (h.vanstel@umcutrecht.nl) Ingrid IE Staal (ingrid.staal@ggdzeeland.nl) Jo MA Hermanns (J.M.A.Hermanns@uva.nl) Augstinus JP Schrijvers (A.J.P.Schrijvers@umcutrecht.nl) Version: 2 Date: 16 January 2012 Author's response to reviews: see over
Response to reviewers Title: Validity and reliability of a structured interview for early detection and risk assessment of parenting and developmental problems in young children: a cross-sectional study Authors: van Stel, Staal, Hermanns, Schrijvers Editor: Please ensure that your manuscript explicitly that ethical approval was granted. o The ethical approval was stated somewhat implicitly. We clarified this in the methods section. o We carefully checked the numbers of the consent and no-consent/non-response groups, and clarified this in the start of the results. Reviewer: Cynthia Leung Internal consistency should be reported o We do not agree that the internal consistency of the SPARK should be reported. The SPARK assessed concerns and perceived need for support from parents on a broad range of topics, and we do not expect that the items are closely related, nor do we try to assess one construct. This inventory is based on a formative measurement model, not a reflective model. The summary score of the SPARK is only intended for scientific purposes, not for use in daily practice. We added a comment on the computation of summary scores in the methods section. The results could have been biased if the nurses had knowledge of the SES and group status of the children and the hypothesis of this study. o The nurses were to some extent aware of the SES of the children, as they asked for the educational level of the parents (a proxy for SES) at the start of the SPARK. Furthermore, they know the neighborhood where the family lives in. They did not have knowledge about the income of the family. In fact, the goal of the structured elaboration before making the final risk assessment is to take account of existing risk factors, including SES. But, most importantly, they were unaware of the study questions of the validation study, including the comparisons between extreme groups and SES groups. This was not clearly stated in the methods section. The goal of the primary study was a comparison of a home visit versus a visit to the well-child clinic. The CHC nurses were not aware of the study goals of the validation study. Therefore we do not think that
the results of the validation study are biased. A better explanation of the study goal is added to the methods section. In the analyses on SES and risk group differences, the median scores of the different groups should be reported. o We included the requested data in the results. More detailed descriptive statistics on the SPARK should be reported. o detailed descriptive statistics are provided in the paper describing the development of the SPARK, which is referred to at several places in the paper. See also our reaction to the comment of the other reviewer. major limitation: the questionnaires used in convergent validation were not validated for use for Dutch families o One of the problems we faced when designing this study was the lack of a gold standard, combined with a lack of validated instrument in this age group. In fact, this lack of available instruments was our main reason to develop and validate the SPARK. We chose to use instruments that cover similar domains as what the SPARK tries to assess, while keeping the burden for parents acceptable. Therefore we chose the ASQ and ASQ:SE, which are indeed not validated in the Netherlands in this age group (but is validated in other countries outside the USA and for older age groups in the Netherlands); the NOSIK, which is the validated translation of the Parenting Stress Index; and the KIPPPI, which is only partially validated but widely used in the Netherlands. We added this to the limitations section. Low correlations should be clarified in the limitations o We have expanded the part in low convergent validity in the limitations. Reviewer: Jane Squires Editing o The manuscript was thorougly checked for mistakes. o We used validity, reliability, and user experience as headings in the results. More information of the SPARK itself o Both reviewers ask for more descriptive information of the SPARK. We intentionally did not include this information, as we did an extensive description in the development paper, to which we refer several times in the paper. We included more information about the SPARK to make the current paper understandable on its own.
For utility, are there any quantitative data? o Yes, and we included these in the results. When is using the SPARK advantageous? I need to be convinced that giving this long interview is advantageous over using separate parent-completed questionnaires o We developed this instrument mainly for use during house calls, to assess the value of house calls for detection of parenting problems. One of the main features of this instrument is the direct interaction between parent and professional: the focus is on interactively discussing with parents the needs of the child and their needs for parenting support. This professional helps the parent with ordering and judging concerns and problems. Whether or not this structured interview is advantageous over using parent-completed questionnaires is a different question that needs to studied. Three arguments are in favour of the SPARK: a) in our current study we observed a response bias, as especially the parents with a child labelled as high risk by the nurse did not return the self-report questionnaires, b) the interview gives nurses the possibility to ask not only about the child, but also about the (functioning of) the family. Nurse reported that especially this part gave them new information relevant for deciding which care and support should be offered, and c) in the Netherlands there is a growing aversion from parents against self-report questionnaires. Parents regard preventive child health care increasingly as a system for detection of child abuse and neglect, instead of as a care provider that supports parents of young children. This threatens the high reach (>95%) that the Dutch system traditionally has between 0-4 years. The non-threatening procedure of the SPARK, and the direct interaction may help in re-establishing the trust of parents in preventive child health care. some talk about the length of time to administer should be included as it is critical for adaptation in many places. o The length of administering the SPARK is about double of the regular time in a wellchild visit. This hampers implementation, in the Netherlands as well as in other countries. Further research is needed (and planned) whether implementing the SPARK is cost-effective. Currently, we are in the process of convincing municipalities (who pay for CHC) and providers of CHC to start implementing the SPARK, based on the results of this and other (soon to be published) studies. Knowledge about costs and benefits seems to be crucial, which is why we are planning a cost-effectiveness analysis of implementing the SPARK. emphasize the preliminary nature of this study (not a stratified sample as it is from one nonrepresentative town, and these are initial data) o The reviewer is not correct in assuming that the sample is from one nonrepresentative town. We collected data on all children born in the whole province of Zeeland, which consists of both towns and villages. This province is representative for a large part of the Netherlands, but less for the highly urbanized areas. The
statement in the discussion was adapted. Furthermore, these are not initial data, as we already reported on another sample of 1140 children in the development paper. This number was added to the introduction.