Methods

Study Design

This study is a retrospective and observational study, because the analyse starts on a pre-collected data from a single Portuguese Hospital scene.

Study Participants

The population considered in this study is all the patients who had entered on the emergency care of Hospital de Santa Maria da Feira, since 1^st October 2005 to 30^th September of 2008 where the Manchester Triage System (MTS) was previously implemented. From all the patients who had gone to this emergency care during the period of time referred, the ones who have more than 16 years old were included in this study (inclusion criteria) which means that they were recruited systematically and registered in a database, representing a total of 336526 cases. Therefore, our sample is our target population. The age 16 was chosen as a criteria because in hospital’ terms people are considered adults since this age. There is no exclusion criteria.

Data Collection methods

This database is a secondary data, since the cases were registered for another purpose than this one. All the information collected on Santa Maria da Feira Hospital was later transferred into an SPSS database.

In order to have a correct performance in the process of evaluation of the waiting times at an hospital’s emergency service, several computer programs had to be used, but one among them was crucial to obtain results and therefore to achieved conclusions: PASW Statistics/ Statistical Package for the Social Sciences - SPSS (a software program used in statistical analysis).

Variables description

The whole data that might be considered on our study derives from the registry of characteristics of all urgency episodes, including many variables likely to be studied and deeply analyzed, that were already present on the SPSS database:

• Sex – nominal variable that was lately codified;
• Date of Birth – numeric variable;
• Date and hour of triage – numeric variable;
• Priority – variable which was lately re-codified into Colours;
• Flowchart outcome – categorical variable;
• Date and hour of the medical observation – numeric variable;
• Result of urgency episode – categorical variable;
• Admission’ Service – categorical variable;
• Date of discharge after admission – numeric variable;
• Date of discharge – numeric variable;
• Result of the Admission – categorical variable;
• Return to ER after 48 hours - categorical and nominal variable;
• Return to ER after 72 hours - categorical and nominal variable.

However, other variables needed to be created and adopted such as:

·         Age – A numerical variable obtained by the difference between the moment they were subjected to MTS and the date of birth;

·         Waiting Times – a numerical variable that informs us about the time that each patient had waited on the emergency room before being seen by a doctor (by the difference of entrance time and medical assessment moment). However, there were some values related to this variable that was unacceptable to sustain objectives, which led to the creation of an error typology. Due to database mistaken values some waiting times were lower than -1 minutes, which is an “error” for this study. Waiting times higher than 24 hours were also rejected, even if this values’ appearance isn’t impossible, it’s highly improbable. 

·         Time spent on ER – numerical variable which results from the difference  between the date of triage and the moment of medical permission to return home, after some or great amount of time on Emergency Room;

·         Exceeded Time - nominal categorical variable which results from the relation between Waiting Times and the time assigned to each colour, due to the priority of patient, reporting if the time was respected or not.

·         Registration of Medical Observation - categorical and nominal variable that informs us if the doctor has registered the moment when he saw the patient;

·         Death – categorical and nominal variable which, from the result of the ER episode, notices if the patient ended up dead or alive.

Planned statistical analysis

Baring in mind that the distribution of the cases for each colour does not follow a normal distribution, in this case the more appropriated measure of frequency is the median.

Using the SPPS, the analyse consisted on estimation of:  the average time from MTS/ triage moment to first medical assessment for each patient comparison; the median waiting time of each colour and comparison with the time that was supposed to be in theory; percentage (%) of patients who had waited more than the specified time for each colour on Manchester Triage System; the mortality rate at every colour (when it respects or exceed the assigned waiting time). Finally we compared the mortality rate and the waiting time and conclude if: the colour was right assessed and if not whether it was related to death; there were other intervenient factors.