High alert drugs screening using gradient boosting classifier

Prescription errors in high alert drugs (HAD), a group of drugs that have a high risk of complications and potential negative consequences, are a major and serious problem in medicine. Standardized hospital interventions, protocols, or guidelines were implemented to reduce the errors but were not found to be highly effective. Machine learning driven clinical decision support systems (CDSS) show a potential solution to address this problem. We developed a HAD screening protocol with a machine learning model using Gradient Boosting Classifier and screening parameters to identify the events of HAD prescription errors from the drug prescriptions of out and inpatients at Maharaj Nakhon Chiang Mai hospital in 2018. The machine learning algorithm was able to screen drug prescription events with a risk of HAD inappropriate use and identify over 98% of actual HAD mismatches in the test set and 99% in the evaluation set. This study demonstrates that machine learning plays an important role and has potential benefit to screen and reduce errors in HAD prescriptions.

Trainee doctor clinics after 24-hour shifts: Effects on patient satisfaction and prescription errors

Introduction: Sleep deprivation impacts clinical performance. However, literature is conflicting, with insufficient focus on patient outcomes. The aim of this study was to assess if patient satisfaction and prescription errors in outpatient clinics were adversely affected when consulting post-call versus non-post-call registrars. Methods: This prospective, quantitative study was set in a large teaching hospital in Singapore. Between November 2015 and February 2016, patients from clinics run by a registrar after 24-hour shift were recruited to post-call group. Patients from non-post-call clinics run by the same registrar were controls. Outcome measures were patient satisfaction, using 5-item 4-point Likert scale questionnaire, and prescribing error rate, defined as number of errors over number of orders. Differences were analysed using chi-squared test. Results: 103 of 106 (97%) patients in 9 post-call clinics and 93 of 105 (90%) patients in 9 non-post-call clinics were recruited. Questionnaire completion rate was 99%. 536 and 526 prescriptions were ordered in post-call and non-post-call groups, respectively. Percentage of top-box responses (greatest satisfaction) was higher in post-call group overall (79.3% versus 62.4%, p<0.001), and for each questionnaire item. There was no significant difference in prescribing errors (1.31% versus 2.28%, p=0.23). Conclusion: Patient satisfaction and prescribing error rates in outpatient clinics were not detrimentally affected. This provides some objective evidence that patients may safely consult post-call registrars. True impacts of sleep deprivation remain poorly understood, and larger, longer term, multicentre studies would inform generalisability. Qualitative studies of fatigue may shed light on complex interactions of emotions that compensate for tiredness.

GOOD PRESCRIPTION: A FIRST STEP TOWARDS THE PREVENTION OF THE MEDICATION ERRORS

STUDY DESIGN: Retrospective analysis of 100 prescriptions of medications. RESULTS: Only 32% prescriptions were had generic names of medications. Right dose and right frequency of medications were missing in 14% and 35% respectively. Ten percent of prescriptions were having illegible handwriting. History of allergy to any was noted in 25% of prescriptions. CONCLUSIONS: Even though Medical Council of India proposed standard prescription format; prescription errors are found to be very common.

278 Duration of Prescription of Dexamethasone in COVID-19 Pneumonitis

Aim The RECOVERY trial showed the effectiveness of using 10 days of dexamethasone for patients hospitalised with Covid-19 disease who required oxygen or mechanical ventilation while an inpatient. This QIP sought to measure compliance with this recommendation. Method Live data was obtained using University Hospital Crosshouse’s electronic prescribing system (JAC). The audit was limited to patients on Ward 2D, which at time of writing was one of the hospitals COVID 19 wards. The first audit was performed on Friday 13/11/20 at 18:00. All dexamethasone prescriptions had the duration of the prescription assessed (at the time all patients were on oxygen therapy for COVID-19 infection). Patients with a course longer than 10 days with explicit documentation to continue past 10 days were counted as a correct prescription in this audit. For the second cycle results were shared with middle tier doctors working on Ward 2D 16/11/20. Notices were also left on all computers on the ward with the message “Please ensure your patient is only prescribed a total of 10 days of dexamethasone unless an extended course is clinically indicated”. A third cycle was carried out 01/12/20. Notices were left on all computers in the combined assessment unit with the same message as above. Results It found most patients were receiving more than the 10 days due to prescription errors (86%). Teaching on the COVID ward resulted in 83% of the prescriptions being correct. Teaching in the receiving unit of the hospital resulted in no change. Conclusions Prescriptions of dexamethasone for Covid-19 patients requiring oxygen on ward 2D were found to be longer than the advised 10 days shown to be effective in the Recovery trial. Teaching on the ward was found to be an effective method to decrease prolonged courses of dexamethasone.

Medication Errors: Reported Prescription Faults and Prescription Error

The present study aimed to evaluate the trends of prescription errors that did not caused any harm to the patients and the prescription errors that were identified before reaching to the patients in the year 2017 at a tertiary care hospital in Kingdom Saudi Arabia. Simple random sampling and sampling based on prescription errors that were identified, documented, and reported before reaching the patients in the first three quarters of 2017 were performed in present observational retrospective study. Descriptive analysis with D’Agostino & Pearson omnibus were applied for normality testing at 95% CI through one-sample t-test to compare the prescription errors that did not cause harm to the patients and were identified before reaching the patient in the first quarter (Q1), the second quarter (Q2), and the third quarter (Q3) of 2017. Total number of prescription errors that did not caused harm to the patients were 1,601 in Quarter 1 further decreased to 1,422 in Quarter 2 and then increased to 1,710 in Quarter 3 of 2017. Furthermore, the total number of prescription errors that did not cause harm to the patients were 1,601 in Quarter 1 further decreased to 1,422 in Quarter 2 and then increased to 1,710 in Quarter 3 of 2017. The current study revealed that prescription errors were common in the tertiary Hospital, Taif, Saudi Arabia. Therefore, educating the prescribers to reduce prescription errors through seminars, conferences, and workshops is essential. Also, a joint training exercise for the pharmacist and doctors would minimize the prescribing errors.

Impact of Electronic Prescription on Prescribing Errors

Aims: There are series of medical errors that can be prevented by taking precautions. Therefore, the study evaluates the impact of the electronic prescribing system on prescription errors. Study Design: A pre-post study design was conducted. Place and Duration of Study: The study was conducted at outpatient pharmacy services of a teaching hospital in Jeddah city. Methodology: Prescriptions were evaluated for the presence of the essential prescription elements such as patient information, drug name, dose, frequency, strength, and other prescription completeness parameters. Results: In the pre-intervention study, 1182 handwritten prescriptions were evaluated, and 6627 errors were detected from these prescriptions. The length of the pre-and post-intervention period was two weeks each. The most prevalent prescribing errors were that of medications written without defined dosage forms were recorded 1653 (55.90%) time followed by prescriptions written by trade names 1493 (22.5%), without route of administration 1266 (19.1%), and without specified duration 1009 (15.2%). However, 1512 prescriptions were evaluated in the post-intervention study, among which 339 errors were detected. The errors included prescriptions written without diagnosis (5.09%), or without doctor’s name or stamp (1.52%), written by trade names (4.49%), without defined dosage forms (4.29%), and without specified duration (2.84%). Conclusion: The study concluded that E-prescribing eliminated prescription errors that resulted from handwritten prescriptions.

Analysis of degree of errors in handwritten medication prescriptions in Rafha, Saudi Arabia

Purpose: To assess the prevalence of handwritten prescription errors in Rafha Central Hospital in Saudi Arabia, and to determine the most predominant type of prescription error. Methods: A cross-sectional study was carried out on randomly selected samples of hand-written prescriptions in out-patient and in-patient pharmacies of Rafha Central Hospital over a five-month period (October 2016 to February 2017). A data collection sheet specially designed for this purpose was used to collect relevant information. The collected prescriptions were analyzed for the presence of prescription errors based on prescription parameters defined by the World Health Organization (WHO) and current guidelines published in British National Formulary (BNF). Descriptive statistics and Microsoft Office were used for processing and analyzing the data collected. Results: Overall, 1019 prescription errors were identified. More than half of the total errors (610; 60 %) were associated with missing patient's information. Moreover, the parameters related to drug and prescriber information were absent in 204 (20 %) and 5 (0.4 %) prescriptions, respectively. In addition, 200 (19 %) miscellaneous errors related to date, legible handwriting and directions for patients were identified. Conclusion: This study discovered errors in hand-written prescriptions. A majority of the prescriptions did not adhere to accepted guidelines. The most common errors are absence of generic names of drugs, non-indication of duration of therapy or prescriber’s contact address, and absence of patient’s weight. Moreover, illegible handwriting was obvious in a substantial number of prescriptions.

Evaluation of the Medication Safety of Chemotherapy Drugs

To evaluate the medication safety of chemotherapy drugs at a tertiary care hospital, with complete reporting of prescription errors, classifying prescription errors, complete detailing of watched medication administration errors (MAEs) by nurses, ordering watched MAEs, and figuring improvement methodologies. Likewise, in relation to side effects, how to overcome side effects, which antiemetic treatments to use, how to survey the appropriateness of requesting and apportioning. An imminent, observational, non-interventional contemplate study was driven at the Oncology Department, Baptist Hospital, Bangalore for half a year. All the data was collected from patient medical records according to case record structure. An aggregate of 70 patients tolerating chemotherapy were observed for information on a sort of side effects, prescription missteps and other relevant information like demographic findings, treatments, and drugs used to manage the adverse effects (AEs) collected from the patient’s medical records. The data was characterized reliant on various parameters. The watched side effects according to different organ frameworks were orchestrated and appeared differently in relation to the distributed writing and bundle embeds. Among the 70 patients, 22 (31.4%) were males and 48 (68.57%) were females. Moreover, the age interval within these two groups was of 20–65. From the 70 patients, the number of chemotherapy cycles was of one for 14 (20%) patients, two for 16 (22.85%), three for 16 (22.85%), four for 5 (7.14%), five for 6 (8.57%), six for 9 (12.85%), and more than six for 4 (5.71%) patients, mostly due to maintenance chemotherapy. The evaluation of our information uncovered that the cancer with the most elevated predominance was breast cancer (24.28%), pursued by blood and bone marrow cancer (5.71%) in females, whereas in males were blood and bone marrow (4.28%), followed by lung cancer (2.85%), non-Hodgkin lymphoma (2.85%), and colon cancer (2.85%). The present study demonstrated that in both gender groups, the most influenced organ framework was gastro intestinal tract (GIT), trailed by skin and subcutaneous tissue, musculoskeletal, blood and nervous system. The most prescribed antiemetic drug was ondansetron (81.42%), and the normally endorsed chemotherapy agents in our setting were shown to be cisplatin (21.42%), carboplatin (17.14%), and paclitaxel (14.28%). The total percentage of errors on the 70 prescriptions was 24.28. Most of the errors were due to drug–drug interactions (10%). The total percentage of errors in drug administration performed by nurses was found to be 11.42%, out of which in 2.85% of the cases, it was used the wrong drug dose. The adverse impacts related with the usage of anticancer medication were surveyed for half a year. The AEs most commonly experienced suggest that for all intents and purposes, all the patients accepting cytotoxic drugs suffered at least one AE. The critical announced MAE rates on our hospital ward (0.04% of medication administration and 0.03% MAE/patient admission) send out an impression of being generally low due to the utilization of current security rules. Emphasize on deep understanding of MAE at individual foundations, is likely going to result in important procedure changes, improved effectiveness of MAE detailing, and various focal points.