Nurse–physician communication and quality of drug use in Swedish nursing homes

Abstract

The objective was to explore the impact of the quality of nurse–physician communication on the quality of psychotropic drug use in Swedish nursing homes, while controlling for resident mix and other nursing home characteristics. Data were collected from a sample of 36 Swedish nursing homes providing care for 1645 residents. Drug use data, along with residents’ demographic characteristics, were obtained from residents’ medication administration lists. Ward nurses reported other residents’ characteristics (e.g. diagnosis and frequency of behavioral problems), and facility characteristics were obtained from head nurses. The quality of drug use was assessed and cross-sectional relationships among study variables were compared. Outcome measures included two drug use quality scores reflecting selection of drug and polymedicine. To assess behavioral problems, a list of the most commonly observed problems was created through a number of steps, including focus groups in the target population. Furthermore, a valid and reliable scale for assessing communication quality was developed. This measure was included in a survey administrated to nurses in the 36 facilities. There was a remarkable variation in the quality of drug use according to the two drug measures. As predicted, the quality of drug use was positively associated with the quality of nurse–physician communication and with regular multidisciplinary team discussions addressing drug therapy and negatively associated with prevalence of behavioral disturbances among residents. Facility size, level of staffing, resident's diagnostic mix, and demographic mix were unrelated to the two drug quality measures. Future efforts to improve the quality of drug use in long term facilities should consider ways of improving communication skills and communication routines among health care professionals.

Introduction

As the elderly population and their average age continue to increase, so does the number of oldest old nursing home residents, who often suffer from a variety of age related diseases and their complications. Alzheimer's disease and other types of dementia can be evidenced by symptoms such as confusion, aggression, self-talk, self-injury, and hallucinations (e.g. Elmståhl, Stenberg, Annerstedt, & Ingvad, 1998). Depression is also a common condition among nursing home residents, resulting in social withdrawal, loss of appetite, and deep sadness (e.g. Heston et al., 1992). The symptoms expressed by persons with dementia are usually a complex function of the disease process and of social and psychological interactions with the environment. It is a challenging task for a nursing home staff to handle the complexity of clinical, social and emotional needs of nursing home residents.

Several types of psychotropic medications, including neuroleptics, sedatives, anxiolytics, and antidepressants, are used to treat a variety of behavioral disturbances. However, while drug treatment may be effective and necessary for some individuals, the risk of further confusion, memory disturbances, and other adverse reactions in this population is substantial. Thus, the clinical literature and current clinical guidelines encourage minimizing the use of inappropriate drugs and careful monitoring of all medications (Avorn & Gurwitz, 1995; Beers & Ouslander, 1989; Harrington, Tompkins, Curtis, & Grant, 1992; Lindley, Tully, Paramsothy, & Tallis, 1992; Medical Products Agency (Sweden) & Medicines Control Authority (Norway), 1995).

The proportion of residents in Swedish nursing homes or similar housing has been increasing, and studies suggest that the proportion of residents with dementia varies between 40% and 80% and that a majority of these residents suffers from confusion (Sandman & Wallblom, 1996). There is also evidence for the fact that the level of drug use in Swedish nursing homes has increased in recent years. In the late 1980s, those living in Swedish nursing homes and sheltered accommodations were prescribed an average of 4–5 drugs each, most to be used on a regular basis (Andersson, 1989). In 1994 the average number of drugs per resident was 7.7, with 80% used on a scheduled basis (National Board of Health and Welfare, 1996). A follow-up study in the same sample of nursing homes showed that the number of medications per resident had increased to 8.7 by 1998 (National Board of Health and Welfare, 1999).

Various attempts have been made to improve drug prescribing practices in nursing homes, including regulations (in the US) and several different educational outreach programs (Avorn et al., 1992; Ray et al., 1993; Rovner, Edelman, Cox, & Shmuely, 1992). Baseline results of a Swedish intervention study (Schmidt, Claesson, Westerholm, Nilsson, & Svarstad, 1998a) designed to improve drug use showed extensive psychotropic prescribing in the 34 nursing homes that participated in the study. After 12 months of multidisciplinary team meetings, involving physicians, nursing staff, and pharmacists, there was a significant decrease in the prescribing of inappropriate psychotropic drugs in the homes assigned to the experimental group. Facilities that received the multidisciplinary team intervention, as well as those with higher staff : resident ratio, were more likely to comply with high-quality drug use criteria for the nursing home population. Through regression analysis, it was determined that resident characteristics (diagnosis, age and sex) had no effect on the quality of drug use at the facility level and that the predictor variables (team interventions and staff : resident ratio) jointly accounted for only 15–20% of the variance among facilities (Schmidt, Claesson, Westerholm, & Svarstad, 1998b).

A number of studies in the US have tried to explore and define the differences in drug-use practices among facilities. Resident demographic and clinical mix are significant predictors of drug use; however, they explain less than 20% of the variation in facility level (Buck, 1988). A few studies have found that nursing home prescribing practices are related to physician factors, such as the number of residents per physician (Ray, Federspiel, & Schaffner, 1980) or frequency of physician contact with residents (Beers et al., 1993). However, individual characteristics such as physician age, gender, years of experience and credentials are unrelated. A few studies find higher rates of drug use in larger facilities and for-profit facilities, but other studies find that facility size and ownership have no effect (Buck, 1988; Garrard, Chen, & Dowd, 1995; Garrard et al., 1991; Ray et al., 1980). Level of staff has also yielded mixed results with a few finding significant effects (Schmidt, Claesson, Westerholm, & Svarstad, 1998b; Shorr, Fought, & Ray, 1994) and others finding a non-significant weak relationship (Buck, 1988; Garrard et al., 1995).

The variation in drug use practices has led to the development of treatment guidelines in order to reduce such differences among facilities. However, it is possible that characteristics of facilities themselves influence the very acceptance of such guidelines (Shorr et al., 1994). As our understanding of different patient outcomes and medical care practice increases, so does the need to understand why such facility differences exist and whether they are medically relevant or not. Past studies in this field typically have been conducted from a medical perspective (which assumes that residents’ clinical characteristics are the most important predictors of variance in prescribing) or from an administrative perspective (which assumes that facility resources and structure are most likely to explain the variation in drug practices). The present study is grounded in a social–psychological perspective, which assumes that professional interaction and communication are also significant determinants of the quality of drug prescribing.