Developing innovative models to translate guidelines into practice.

A study described in this issue explored the feasibility of implementing two specific clinical practice guidelines. Using a low-cost intervention, Resnick et al. improved medical record documentation of care recommended in pain and falls practice guidelines. The intervention involved providing written materials to volunteer nursing home (NH) staff. One might hypothesize that an approach that supplemented this education with follow-up consultation might have been even more effective, although also more expensive. One NH staff training intervention model, which allows frequent and routine consultation to be provided about practice guideline implementation in a cost-effective manner, is described later in this editorial. One of the strengths of this study is that Resnick et al. focused on specific processes of care as the primary measure of how well guidelines were implemented. These care processes were measured by independent research staff using standardized chart review protocols. This direct measurement of care processes has important advantages over evaluation measures that involve NH staff self-reports of outcomes or changes in policy. In the case of fall and pain outcome measures, one particularly must be concerned about how NH staff defines these conditions. In addition, long periods of monitoring are necessary to detect significant improvements in low incidence conditions such as falls. Alternatively, NH staff reports of policy changes regarding assessment and treatment of these conditions are several steps removed from improved medical record documentation or, more importantly, the actual care received by residents. The care process measures collected by Resnick et al. might be criticized by the fact that they are derived primarily from medical record data. Several studies have demonstrated that chart documentation of care processes related to incontinence, pressure ulcers, restraints, and nutrition do not translate into better direct care as determined by resident interview or direct observation. Despite this concern, one can still plausibly argue that the more specific chart documentation that was measured by Resnick et al. (eg, better use of standardized fall or pain assessments) is an important step to improving care even if there is not evidence that the actual care received by residents changed. However, future efforts to translate guidelines into practice should consider evaluating both the quality of chart documentation and care actually received by residents independent of NH staff-generated reports. In regard to such future research, it would not be surprising if ongoing support by consultants with expertise in the practice guidelines being implemented will be necessary to improve the actual care provided to residents. Such ongoing support has clearly been necessary in efforts to implement urinary incontinence practice guidelines in NHs. Unfortunately, support by clinical experts past a startup stage is expensive and could be cost-prohibitive if it requires frequent travel to the NH. In consideration of these cost issues, new models to both implement guidelines and provide ongoing consultation should be considered. These models should exploit information technologies that are slowly being adapted to the NH setting. To illustrate how such a consultative model might be designed, consider preliminary work being conducted in the area of nutritional care. A software program has been designed that operationalizes nutritional care practice guideline assessment and treatment recommendations. Quality monitoring protocols have also been developed that allow NH supervisory and consultative research staff to monitor the actual care processes provided to residents through direct observations (eg, quality of feeding assistance). These quality monitoring data are also integrated into the software, which generates reports consistent with federal documentation requirements for residents at risk for weight loss. A research or survey team that conducted chart reviews in a NH that used this software could easily locate clear documentation that most nutritional practice guideline recommendations are being implemented appropriately for at-risk residents. More importantly, consultative staff with expertise in nutritional care can monitor all data entered into the software from an off-site location. The same quality-monitoring reports documenting improved care provision that are generated by NH staff then can be generated by consultative staff. Common access to these reports facilitates meetings between NH and consultative staff that can be efficiently conducted by either telephone or even e-mail. In cases in which problems in the implementation of nutritional care processes are revealed by the quality-monitoring protocol, brief video vignettes are being developed to allow timely training (eg, employee can be asked to watch a 5-minute video to show how they should socially interact with residents while providing feeding assistance). Address correspondence to John F. Schnelle, PhD, Director, University of California–Los Angeles Borun Center for Gerontological Research, 7150 Tampa Ave., Reseda, CA 91335. E-mail: [email protected].