Journal of Public Health Advance Access originally published online on May 19, 2007
Journal of Public Health 2007 29(3):321; doi:10.1093/pubmed/fdm021
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Dealing with incomplete and inaccurate data in public health: case study of a health equity audit of health visiting services
Lincoln A. Sargeant
Department of Public Health and Primary Care
University of Cambridge
Forvie Site, Robinson Way
Cambridge CB2 0SR, UK
Nina Heaps
Penny Miller
Cambridgeshire Primary Care Trust
East Cambridgeshire and Fenland Locality
Central Hall, 52-54 Market Street
Ely CB7 4LS, UK
E-mail: ls348{at}medschl.cam.ac.uk
Public health practitioners often make decisions with incomplete or inaccurate data. A health equity audit begins with agreement of priorities with relevant partners and the collection and analysis of data to establish the equity profile at baseline. In order for all partners to accept the findings of an equity profile and use them as the basis for further action to correct inequity of service provision, there needs to be agreement on the suitability of the data sources underpinning the equity profile. Sensitivity analysis is an approach that may resolve these challenges, and its use is illustrated in a health equity audit of health visiting services.
There was consensus that data on 15 tasks were important for adequately describing the workload of health visitors, and indirectly the health care needs of children aged 0–4 years. Caseload weighting was done with each health visitor providing details of their caseload. Weights were applied to reflect the time differences in managing cases and used to estimate the staff needed to cover the workload. The difference in staff capacity was reflected in an equity score. Sensitivity analysis was done to determine how robust the score was to changes in specific data used. The core programme in some areas involved fewer visits to each child. Unequal weights were given to each child on the caseload to reflect this difference.
The most complete caseload represented 39% of the population of eligible children. The most important contributor to workload was the number of children on the caseload accounting for 50%. The equity score was sensitive to three factors: differential weighting for the number of children aged 0–4 years on the caseload, the number of ‘medium dependency’ families and the prevalence of postnatal depression. ‘Medium dependency families’ were those requiring additional support beyond the core programme of visits and interventions offered to each child, but there was no common definition across all teams.
Sensitivity analysis may be used to determine the relative importance of various data sources and to gauge the potential impact of incomplete or inaccurate data on public health decision-making. One practical implication of this finding was that the data collection could be targeted for future cycles of the health equity audit and the process of doing an equity profile could be greatly simplified. Since about half of the workload was due to the size of the caseload, comparison of the average caseload per health visitor across teams could be used to monitor equity. This could be supplemented with data identifying children and families ‘at risk’.1,2
Another important outcome of the sensitivity analysis was the demonstration that differences in practice could account for inequity and that equity could be achieved without the need for additional investment. Finally, by identifying ‘medium dependency’ families and postnatal depression as tasks that strongly influence the equity profile, sensitivity analysis highlighted the need to ensure that these data sources were robust.
 |
Conflict of interest statement |
|---|
 TOP Conflict of interest statement References |
|---|
None declared.
|
References |
|---|
|
TOPConflict of interest statementReferences |
|---|
- Steel N, Reading R, Allen C. An assessment of need for health visiting in general practice populations. J Public Health Med (2001) 23(2):121–8.
[Abstract/Free Full Text] - Crofts DJ, Bowns IR, Williams TS, et al. Hitting the target: the equitable distribution of health visitors across caseloads. J Public Health Med (2000) 22(3):295–301.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||