Objective: Oral administration of hypnotics is one of the many causes of falls. Recently, non-benzodiazepine sedative hypnotics are frequently prescribed in high-risk patients of falls, because these hypnotics may avoid muscle relaxant action. The objective of this study was to determine which hypnotic has a lower risk for fall-related accident. Methods: We examined relationship between the hypnotics prescribed and incident reports of falls in hospitalized patients, retrospectively. Results: We studied 318 cases of falls in our hospital from April 2015 to March 2016. During the period, hypnotics frequently prescribed on demand were as follows; zolpidem: 2,614 cases, eszopiclone: 646 cases, brotizoram: 571 cases. The cases that had a hypnotic within 24 hours before fall are zolpidem 9 cases (0.34% of total zolpidem on demand administration), eszopiclone 1 case (0.15%), and brotizoram 3 cases (0.53%). Conclusions: As previously reported, falls related hypnotics were more frequent with benzodiazepine sedative hypnotics than non-benzodiazepine. Especially, patients who had administrated eszopiclone as a single use rarely experienced falls. Our findings indicate that eszopiclone is estimated to be a safer hypnotic than other hypnotic for the treatment of insomnia in high-risk patients of falls.

View full abstract

PURPOSE: Reading a prescription incorrectly may cause serious accidents in some cases. People who experience the same pattern repeatedly create a fixed pattern in their minds. It is possible that such fixations, formed from prior empirical contexts, contribute to reading errors. The author demonstrated this hypothetical cause of reading errors in two experiments. METHOD: In two experiments with two conditions each, the author asked the participants to transcribe 10 handwritten percentage digits. The seventh digit was the target digit and was the same for both conditions in both experiments. For experiment 1, the author prepared a double digits condition consisting of all double digits, except the target, and a single digits condition consisting of all single digits. For experiment 2, he prepared two conditions, in which single and double digits were mixed half-and-half. The difference between the conditions was the previous digit of the target. A single digit preceded the target in one condition and a double digit preceded the target in the other condition. RESULTS: In the experiments, the participants tended to misread the target as “50%” although it was “5%” in reality. Because the left-hand “o” of the “%” was large, they mistook it as zero (0). More participants in the double digits condition misread the target than those in the single digit condition in the experiment 1, and did than those in the each mix digits condition of the experiment 2. CONCLUSION: The results indicate that not only misleading objects but also the fixation formed from prior empirical context raise the rate of occurrence of misreading.

View full abstract

Objective: This study aimed to develop a monitoring procedure for case detection based on integrating adverse event data extracted from medical fee information and other data. Cases of bone fractures and intracranial bleeding due to falls were detected in this study. Materials and Methods: The following three procedures for case detection were attempted among 64,832 patients admitted to our hospital from April, 2012 and discharged until December, 2015: first procedure: case detection based only on medical fee information in which cases of patients with bone fractures and intracranial bleeding were collected from the day after admission; second procedure: case detection using review of medical records of cases collected by the first procedure; third procedure: case detection based on integrating the data obtained by the second procedure and the data on reported incident cases. Result: He numbers of cases of patients with bone fractures and intracranial bleeding detected by each procedure were 313 and 324, respectively, for the first procedure, 21 and 11 for the second procedure, and 24 and 11 for the third procedure. Assuming that all actual cases are detected using the third procedure, the first procedure had 87.5％ sensitivity, 99.5％ specificity and 6.7％ positive predictive value for detecting cases of bone fractures, and 100%, 99.5% and 3.4%, respectively, for detecting cases of intracranial bleeding; the second procedure using additional review of medical records had 100% specificity and 100% positive predictive value for detecting both cases of bone fractures and intracranial bleeding. Conclusion: A procedure for case detection based on integrating the data on adverse events detected by review of medical records of cases narrowed down by medical fee information and the data on reported incidents can be applied to monitoring of adverse events.

View full abstract