Table 2 Predictors of subsequent events
From: Using machine learning to predict subsequent events after EMS non-conveyance decisions
The patient had subsequent event and the model predicted there will be one (n = 17) | ||
|---|---|---|
The study group agree with model there will be subsequent event (n = 17) | ||
Signs/symptoms | n | % |
Musculoskeletal symptoms | 6 | 35.3 |
Infection | 3 | 17.6 |
Non-specific complaints | 2 | 11.8 |
Abdominal pain | 2 | 11.8 |
High blood pressure | 1 | 5.9 |
Ear pain | 1 | 5.9 |
Nasal bleeding | 1 | 5.9 |
Frequent caller with minor problem | 1 | 5.9 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 76.5% of the cases (13/17) | ||
The patient had subsequent event, but the model did not predict one (n = 20) | ||
|---|---|---|
The study group agree with model there won't be subsequent event (n = 7) | ||
Signs/symptoms | n | % |
Frequent caller with minor problem | 3 | 15.0 |
Fall due to alcohol misuse | 1 | 5.0 |
Urinary catheter blocked | 1 | 5.0 |
Psychological symptom | 1 | 5.0 |
Allergic reaction | 1 | 5.0 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 0% of the cases (0/7) | ||
The study group disagree with model there won't be subsequent event (n = 13) | ||
|---|---|---|
Signs/symptoms | n | % |
Psychological symptom | 5 | 25.0 |
Musculoskeletal symptoms | 3 | 15.0 |
Non-specific complaints | 2 | 10.0 |
Convulsion | 1 | 5.0 |
Violence/assault | 1 | 5.0 |
Nausea | 1 | 5.0 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 76.9% of the cases (10/13) | ||
The patients didn’t have subsequent event, but the model predicted that there will be one (n = 20) | ||
|---|---|---|
The study group agree with model there will be subsequent event (n = 18) | ||
Signs/symptoms | n | % |
Musculoskeletal symptoms | 11 | 55.0 |
Infection | 2 | 10.0 |
Breathing trouble | 2 | 10.0 |
Psychological symptom | 1 | 5.0 |
High blood pressure | 1 | 5.0 |
Diarrhoea | 1 | 5.0 |
The agreement with the patient/guidance to primary health care or ED next or following days 66.7% of the cases (12/18) | ||
The study group disagree with model that there will be subsequent event (n = 2) | ||
|---|---|---|
Signs/symptoms | n | % |
Musculoskeletal symptoms | 1 | 5.0 |
Breathing trouble | 1 | 5.0 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 0% of the cases (0/2) | ||
The patients didn’t have subsequent event and the model did not predict one (n = 20) | ||
|---|---|---|
The study group agree with model there won’t be subsequent event (n = 15) | ||
Signs/symptoms | n | % |
Psychological symptom | 4 | 20.0 |
Chest pain | 3 | 15.0 |
Abdominal pain | 2 | 10.0 |
Fall (elderly with minor conditions) | 2 | 10.0 |
Musculoskeletal symptoms | 1 | 5.0 |
Fall due to alcohol misuse | 1 | 5.0 |
Toxic effect of carbon monoxide | 1 | 5.0 |
Bicycle accident | 1 | 5.0 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 0% of the cases (0/15) | ||
The study group disagree with model there won’t be subsequent event (n = 5) | ||
|---|---|---|
Signs/symptoms | n | % |
Non-specific complaints | 2 | 10.0 |
Psychological symptom | 1 | 5.0 |
Musculoskeletal symptoms | 1 | 5.0 |
Symptoms of apoplexia | 1 | 5.0 |
The agreement with the patient/guidance to visit primary health care or ED next or following days 60.0% of the cases (3/5) | ||