Predicting Child Risk for Distress With a Painful Procedure

Purpose:

The problem is that clinicians need decision support tools to identify child risk for distress with procedural pain in order to allocate resources and provide interventions that are feasible to implement in a clinical setting. The purpose of this study was to develop a web-based computer application (app) that assists providers in identifying families that are able to provide distraction (i.e. decision support) and those that need additional support from a professional. For those parents able to provide support, an app that provides tailored instructions and training for parents to function as distraction coaches during their child’s IV insertions is a valuable tool. The research-based app, called Children-Parents and Distraction (C-PaD), was developed from research and has been validated, but needs feasibility testing in clinical settings.

Most young children experience a number of painful medical procedures, such as immunizations, as part of their routine health maintenance. Many children with health problems require additional painful procedures such as venipuncture and IV insertions related to diagnostic testing and disease treatment. Inadequate pain management related to needle sticks can create anxiety in the child, behavioral distress, and may have negative long-term consequences (Kennedy, Luhmann, & Zempsky, 2008; Taddio et al., 2012).

Despite research to support intervention for acute procedure related pain in children, it remains undertreated (Stevens et al., 2011; Stinson, Yamada, Dickson, Lamba, & Stevens, 2008). Topical anesthetics (TA) are effective at reducing needle pain in most, but not all children (Kleiber, Schutte, et al., 2007). Other interventions that work with TA are needed. Distraction is a relatively simple and effective cognitive behavioral intervention for reducing pain and distress for young children undergoing painful procedures (Koller & Goldman, 2012; McCarthy et al., 2010; Stinson et al., 2008; Uman et al., 2013). Parents typically want to help their young child during medical procedures and many can be trained to be distraction coaches for their children (Kleiber, McCarthy, Hanrahan, Myers, & Weathers, 2007; Power, Liossi, & Franck, 2007). However, about 15% of children do not respond to the distraction efforts of their parents, and display high levels of distress behavior (McCarthy et al., 2010). These children need professionals to provide interventions.

Methods:

Predictive models of distress used in the C-Pad prototype were developed from a multisite study of 542 children and parents using data mining techniques. Predictive data mining uses specific cases with known outcomes to construct models that use independent variables (such as demographics and survey responses) to predict outcomes for future cases. Generalization, the performance of the model on unseen data, was further improved by an automatic feature selection method. Using this method, individual items are automatically selected from the data to build the best models for predicting outcomes. Data from a second multisite study (n =574) were used to update the predictive models in the C-Pad, using the same data mining techniques.

Before translating the C-PaD app for use in the community, the prototype needed: a) a new interface to include a responsive design for use on multiple devices (e.g. tablets and smart phones), and b) revised training components including new graphic design and an updated teaching video. From an Information Technology perspective this involves: Web application development, beta testing and web service development to create a sustainable product.

Results:

The C-PaD app has gone through several levels of testing before being released to production. The individual developer performed unit testing throughout the development process. The developers performed additional system tests in this environment and then turned over to a user group to perform user acceptance/beta testing. When all issues were addressed the application was then moved to the production environment. The pilot version was released to a small group of end users for pilot testing in “real world” conditions before final release of the application to the target user community. Feasibility testing was conducted in the Emergency Department and Pediatrics Specialty Clinic Phlebotomy Lab with parents and healthcare providers completing an evaluation on app device and participated in a brief interview. The C-PaD app will be demonstrated and ongoing results and modifications from feasibility testing will be presented.

Conclusions: Use of the C-PaD for decision support is expected to decrease child pain, anxiety and distress with venipunctures, increase parent participation and decision making in providing distraction during IV insertions, and increase provider awareness of individual child risk for distress in order to increase procedural success and appropriate allocation of resources. Next, clinical translation research to determine the effectiveness of the C-PaD in routine practice and translation to community settings is needed.