Relationships between Ambient Sounds and the Accuracy of Pharmacists' Prescription-Filling Performance
Flynn, Elizabeth Allan, Barker, Kenneth N., Gibson, J. Tyrone, Pearson, Robert E., Smith, Leo A., Berger, Bruce A., Human Factors
Consumption of accurately filled prescriptions is an important component of patient recovery. However, evidence shows that there may be a problem with the accuracy of prescription filling. Medication dispensing error rates ranging from 2% to 24% have been detected by visual inspection in pharmacies (Allan, 1994; Allan, Barker, Malloy, & Heller, 1995; Buchanan, Barker, Gibson, Jiang, & Pearson, 1991; Guernsey et al., 1983; Kistner, Keith, Sergeant, & Hokanson, 1994; McGhan, Smith, & Adams, 1983). For example, if 5% of the two billion prescriptions filled in the United States each year are filled in error, 100 million errors occur annually (National Association of Chain Drug Stores, 1994).
A limited number of factors have been associated with medication dispensing errors. These include excessive prescription workload (Allan, 1994; Buchanan, 1989; Guernsey et al., 1983), insufficient illumination level (Buchanan et al., 1991), interruptions (Allan, 1994), and distractions (Allan, 1994). A review of the human factors literature was performed to identify additional variables in the work environment that may be related to dispensing errors.
Ambient sound was identified as a variable that has had a significant effect on the accuracy of human performance. A number of reviews of the effects of ambient sound characteristics on performance are available (Cohen & Weinstein, 1981; Gawron, 1982; Glass & Singer, 1973; Glass, Singer, & Pennebaker, 1977; Kjellberg, 1990; Kryter, 1985; Sundstrom, 1986). Three theories propose ways in which noise affects human performance (Cohen & Weinstein, 1981). Noise can focus attention by increasing arousal, but the level of performance depends on the complexity of the task (Broadbent, 1971). Cohen (1978) proposed that performance is affected by the predictability and controllability of the noise. Poulton (1978, 1979) theorized that adverse effects are attributable to distraction that results from the initiation of intermittent noise, which may mask environmental cues and internal speech. Jones and Broadbent (1991) summarized research on the effects of noise on performance. The list that follows is drawn from this work (limited to those stimuli that may be present in a pharmacy):
1. Bursts of noise decrease performance when the participant is (simultaneously) receiving information.
2. Frequent noises are distracting at first but start having an arousal effect after a long duration of exposure (length not specified).
3. The perceived level of controllability over noise affects performance; decreased control is associated with poorer performance.
4. Performance of simple tasks is not affected by continuous noise.
5. Performance of more than one task simultaneously results in diminished performance of one of the tasks in the presence of noise.
6. Intelligible speech and irrelevant speech disrupt performance.
This summary, along with consideration of research by Glass and Singer (1973), suggests that uncontrollable noise and unpredictable noise might be associated with an increased frequency of dispensing errors. Pharmacists are exposed to uncontrollable and unpredictable noise as they fill prescriptions, including loudspeaker announcements, telephone rings, and conversations. A study of the effect of ambient sound on the accuracy of pharmacists' prescription filling may provide evidence of whether these factors adversely affect pharmacists to the point that they commit errors.
Another characteristic of ambient sound that may affect the quality of performance is loudness. Studies of performance on tasks while exposed to decibel (dBA) levels similar to what might be found in a pharmacy (less than 80 dBA) have conflicting conclusions. Gawron (1982) studied the effects of 55, 70, and 85 dBA noise on the performance of a tracking task with two levels of task complexity. Accuracy improved in a linear manner on both simple and complex tasks as loudness increased. …