Using data mining to predict emergency department length of stay greater than 4 hours: Derivation and single-site validation of a decision tree algorithm

OBJECTIVES

Health services have an imperative to reduce prolonged patient length of stay (LOS) in ED. Our objective is to develop and validate an accurate prediction model for patient LOS in ED greater than 4 hours using a data mining technique.

METHODS

Data were collected from a regional Australian public hospital for all ED presentations between 1 January 2016 and 31 December 2017. A decision tree algorithm was built to predict patients with an ED LOS >4 hours. A total of 33 attributes were analysed. The performance of the final model was internally validated. Clinically relevant patterns from the model were analysed.

RESULTS

The accuracy of the model was 85%. We identified that patients at our site who were at high risk of ED LOS >4 hours were those who were waiting in ED for a medical consultation, or those who were waiting for a urology, surgical, orthopaedic or paediatric consultation if the request for consultation occurred more than 2 hours after the patient was first seen by an ED doctor.

CONCLUSION

This model performed very well in predicting ED LOS >4 hours for each individual patient and demonstrated a number of clinically relevant patterns. Identifying patterns that influence ED LOS is important for health managers in order to develop and implement interventions targeted at those clinical scenarios. Future work should look at the utility of displaying individual patient risk of ED LOS >4 hours using this model in real-time at the point-of-care.

Fetal folate C677T methylenetetrahydrofolate reductase gene polymorphism and low birth weight

The aim of this study was to determine if fetal C677T methylenetetrahydrofolate reductase (MTHFR) genotype contributes to low birth weight. The study group consisted of 243 term babies with a birth weight<10th centile for gestational age, with subgroup analyses for those <1st centile. The control group consisted of 132 term babies with a birth weight 3.3-3.8 kg. Odds ratio analyses with 95% confidence intervals (CI) were calculated for carriage of the t allele and overall genotype frequencies. There was no significant difference in carriage of the t allele between study and control groups, odds ratio (OR) 0.79 (95% CI, 0.57-1.09). No differences were observed for frequencies of heterozygote and recessive homozygote genotypes for the two populations. In the subgroup analyses, no statistical differences were observed in the t allele frequency, frequency of the heterozygote or homozygote genotype. Trends were seen and the study suggests that fetal C677T MTHFR genotype may be a factor contributing to birth weight. The potential may exist to influence clinical outcome by maternal folate supplementation.

Management of a pregnancy complicated by type III spinal muscular atrophy

The authors report the successful management of a pregnancy in a patient with spinal muscular atrophy (SMA) type III. It is a genetically inherited condition causing increasing weakness of the skeletal muscle. The patient in our case was confined to a wheelchair due to marked weakness in her lower limbs. A review of the available literature identified potential risk factors for the antenatal, intrapartum and postpartum period. These include increased risk of thromboembolism, urinary tract infections, intrauterine growth restriction and preterm delivery, anaesthetic problems and increased risk of uterine atony with subsequent postpartum haemorrhage. The authors report the management of these risk factors and the reasons for delivery by Caesarean section at 32 weeks. Apart from a decline in muscle function postdelivery requiring physiotherapy, there were no adverse outcomes for mother or baby. In both this case and on review of the literature, it is proven that a successful pregnancy is possible with SMA.

Doctor-initiated consultations: a study of communication between general practitioners and patients about the need for reattendance

It has been suggested that general practitioners have the potential to regulate a large percentage of their workload through their control of 'doctor-initiated' consultations. A survey was made of 300 consecutive consultations in a group practice. After their consultation patients completed a questionnaire asking what advice the doctor had given them on the need to reattend. At the same time the general practitioner completed a similar questionnaire about the need for reattendance and the advice given. The general practitioners judged that 74% of patients definitely or possibly needed to reattend, and only 26% definitely did not need to reattend. The coefficient of agreement between patients' and doctors' views on whether reattendance had been recommended was only 0.41. Thus the room for control of doctor-initiated consultations is limited by both clinical considerations and the apparent difficulty of accurately communicating the doctor's advice on reattendance to the patient.