Adverse Event Signal Detection Using Patients' Concerns in Pharmaceutical Care Records: Evaluation of Deep Learning Models.

BACKGROUND: Early detection of adverse events and their management are crucial to improving anticancer treatment outcomes, and listening to patients' subjective opinions (patients' voices) can make a major contribution to improving safety management. Recent progress in deep learning technologies has enabled various new approaches for the evaluation of safety-related events based on patient-generated text data, but few studies have focused on the improvement of real-time safety monitoring for individual patients. In addition, no study has yet been performed to validate deep learning models for screening patients' narratives for clinically important adverse event signals that require medical intervention. In our previous work, novel deep learning models have been developed to detect adverse event signals for hand-foot syndrome or adverse events limiting patients' daily lives from the authored narratives of patients with cancer, aiming ultimately to use them as safety monitoring support tools for individual patients. OBJECTIVE: This study was designed to evaluate whether our deep learning models can screen clinically important adverse event signals that require intervention by health care professionals. The applicability of our deep learning models to data on patients' concerns at pharmacies was also assessed. METHODS: Pharmaceutical care records at community pharmacies were used for the evaluation of our deep learning models. The records followed the SOAP format, consisting of subjective (S), objective (O), assessment (A), and plan (P) columns. Because of the unique combination of patients' concerns in the S column and the professional records of the pharmacists, this was considered a suitable data for the present purpose. Our deep learning models were applied to the S records of patients with cancer, and the extracted adverse event signals were assessed in relation to medical actions and prescribed drugs. RESULTS: From 30,784 S records of 2479 patients with at least 1 prescription of anticancer drugs, our deep learning models extracted true adverse event signals with more than 80% accuracy for both hand-foot syndrome (n=152, 91%) and adverse events limiting patients' daily lives (n=157, 80.1%). The deep learning models were also able to screen adverse event signals that require medical intervention by health care providers. The extracted adverse event signals could reflect the side effects of anticancer drugs used by the patients based on analysis of prescribed anticancer drugs. "Pain or numbness" (n=57, 36.3%), "fever" (n=46, 29.3%), and "nausea" (n=40, 25.5%) were common symptoms out of the true adverse event signals identified by the model for adverse events limiting patients' daily lives. CONCLUSIONS: Our deep learning models were able to screen clinically important adverse event signals that require intervention for symptoms. It was also confirmed that these deep learning models could be applied to patients' subjective information recorded in pharmaceutical care records accumulated during pharmacists' daily work.

[The Usefulness of a Pamphlet for Preventing Exposure to Antineoplastic Agents at Cancer Patients' Homes and an Awareness Survey for Pharmacists about Providing Information to Cancer Patients about Exposure to Antineoplastic Agents].

As the number of patients undergoing outpatient chemotherapy has increased, there is concern that cancer patients' family members are unknowingly exposed to antineoplastic agents at home through cancer patients' excrement or other secreted materials. In this study, we created a pamphlet that introduces several methods to prevent exposure to antineoplastic agents at home and conducted a questionnaire survey to assess the usefulness of the pamphlet. The results indicated that more than 90% of patients believed that the pamphlet was "useful" or "very useful" for ensuring safety with respect to antineoplastic agents at home. Further, most patients responded that the pamphlet decreased their anxieties about their disease and/or treatment. In order to examine pharmacists' involvement in providing information to cancer patients about exposure to antineoplastic agents, we conducted another questionnaire survey, with pharmacists working at Sapporo-Higashi Tokushukai Hospital and Sapporo Tokushukai Hospital. The results indicated that 41 out of 46 pharmacists practiced medication counseling; however, 39 pharmacists did not provide patients with instructions on ways to prevent exposure to antineoplastic agents at home. Their primary reason was a lack of adequate information to do so. Accordingly, the pamphlet prepared in our study would be an effective way to provide guidance for preventing exposure to antineoplastic agents at home.

Combined low calcium and lack magnesium is a risk factor for motor deficit in mice.

The populations of the Kii Peninsula in Japan and of Guam present high incidences of amyotrophic lateral sclerosis and Parkinsonism-dementia complex. It is thought that low levels of calcium (Ca) and magnesium (Mg) in the drinking water are involved in the pathogenesis of these diseases. The present study aimed to test the hypothesis that catalepsy, behavioral immobility and a Parkinsonian symptom results from functionally impaired dopaminergic neurons in mice fed low amounts of Ca and Mg (LCa/Mg). A group of mice fed a LCa/Mg diet for 6 weeks was compared to a control group on a standard diet. Cataleptic symptoms such as akinesia and rigidity were measured by the bar test. The anti-parkinsonian drugs dopamine (DA) precursor L-3,4-dihydroxy phenylamine (L-DOPA), the selective DA receptor D(2) agonist bromocriptine, and the DA releaser amantadine were tested for their effects on induced catalepsy. The mice developed catalepsy after 3 weeks on the LCa/Mg diet. LCa/Mg diet-induced catalepsy was improved by the administration of L-DOPA (50-200 mg/kg i.p.) in combination with benserazide (25 mg/kg i.p.), or of bromocriptine (0.25-4 mg/kg i.p.) or of amantadine (5-20 mg/kg i.p.). Immunohistochemical staining revealed that the intensity of tyrosine hydroxylase fluorescence was significantly decreased in the substantia nigra at the 6th week of LCa/Mg feeding in comparison with pair-fed controls. These results suggest that catalepsy in LCa/Mg mice results from hypofunction of the dopaminergic neurons. Moreover, our results support the hypothesis that LCa/Mg intake is one etiological factor in neurodegenerative disorders, including Parkinson's disease.

[Prediction of gemcitabine treatment discontinuation based on platelet count before first administration].

Decrease in white blood cell (WBC), neutrophil or platelet (PLT) count due to treatment with gemcitabine (GEM) is a dose-limited factor (DLF). Even for cases that satisfy the standard criteria for initiation of GEM therapy, the scheduled therapy is reportedly occasionally discontinued because of decreased PLT or WBC count. Here, a retrospective study was made to predict the factor causing discontinuation of GEM treatment before its first administration. The results showed that PLT count immediately before the first administration was significantly less in the unfinished administration group than in the finished group. It was also demonstrated that a PLT count less than 16×10/4 mL before the first administration of GEM was the significant risk factor leading to discontinuation of GEM treatment. Thus, it was suggested that this result would be available as the dose reduction standard to determine the first dose of GEM treatment.

[Investigation of factors associated with time change in breath alcohol concentration after paclitaxel administration].

Paclitaxel (PTX) injection is presently used for various types of cancer therapy. Because PTX is an insoluble medicine, alcohol is included. Therefore, after PTX administration, the breath alcohol concentration (BAC) detected is reported. We measured BAC after PTX administration and conducted an investigation into change over time. We also investigated factors associated with BAC. As a result, the BAC rate of detection just after the PTX administration was 52%. BAC was confirmed 3 hours after the administration. Therefore, sufficient rest periods were necessary after PTX administration and the need to avoid driving was proved. From comparison of the BAC detected and undetected groups, no significant sex differences were found in the BAC detected group in age, dosage, or body surface area. However, about a dosage per the time, many other things were significant in the BAC detected group in terms of dosage per time. From these results, it was suggested that the BAC detection rate rose with the dosage per time. However, one cannot predict BAC detection by the usual medical examination. Thus, when receiving PTX treatment on an outpatient basis, the patient should be warned not to drive to and from the clinic.

Alterations in cognitive function in prepubertal mice with protein malnutrition: relationship to changes in choline acetyltransferase.

We have found that protein malnutrition (PM) causes a significant impairment of memory-related behavior on the 15th and 20th day after the start of PM (5% casein) feeding in prepubertal mice but not in postpubertal mice, as measured by a passive-avoidance task. This impairment was almost completely reversed by merely switching to a standard protein (20% casein) diet on the 10th day after the start of PM. However, the reversal was not observed when the switching to a standard protein regimen was done on the 15th day of the PM diet. Interestingly, the impairment of memory-related behavior on the 20th day was improved by the chronic administration of physostigmine (0.1 mg/kg/day x last 10 days, i.p.), a cholinesterase inhibitor. To correlate brain cholinergic neuron function with the memory-related behavior impairment induced by PM, microphotometry was used to determine the histological distribution of the imunofluorescence intensity for choline acetyltransferase (ChAT), a functional marker of presynapse in cholinergic neurons. The change in the intensity of fluorescence indicated that ChAT protein was decreased in the hippocampus (CA1, CA3 and dentate gyrus) on the 20th day after PM feeding in comparison with controls. These results suggest the possibility that the memory-related behavior deficits observed in prepubertal mice with PM are caused by a dysfunction of the cholinergic neurons in the hippocampus.