Results 

General Statistics

The total number of cases present in the sample was 336526, in which 42,1% corresponded to male cases.

The next figure (figure 3) expresses the frequency of missing cases, which correspond to the cases catalogued with “without colour”, whose won’t be taken into account in the following analysis. The “valid” cases would be the ones catalogued with the colours: Red, Orange, Yellow, Green, Blue and White, consisting in 336477 cases available to further statistical procedures.

Following the objectives we proposed ourselves to accomplish, appears the need to confirm the correspondence between the colour assigned and the waiting time, if the mean time from MTS to first medical assessment determined in theory is being done successfully in practice. And, in the cases where patients waited more than expected, analyze the outcome of those patients.

So, talking about the standard values for each colour, it becomes possible to compare these ones to the values we verified after analyzing the database (with creation of variable “waiting time” and estimation of its median):

Looking at the chart on the distribution of waiting times by different colors, comparing it with the time suggested by MTS, multiple results can be advanced, namely: in the case of patients who were assigned the color Red, they waited on average about 11 minutes, when it would be expected a value close to 0 minutes, as suggested in the MTS.

Something similar happens in the Orange’s cases where there is a median of 15 minutes, when the suggested time was 10 minutes. In the other colours the median value is lower than suggested by MTS and medians were respectively 37 minutes for the Yellow colour, 49 minutes for the Green colour and 63 minutes for the Blue cases.

It is important to note that the values reported above show a central tendency, so if for example a patient who has been assigned the color blue in median waited about 63 minutes, it just means that 50% of patients waited more than 63 minutes and 50% waited less. For example for the blue case, about 2.5% (97.5th percentile) of patients waited (on average) about 316 minutes. The same reasoning can be applied to the other colors.

Some cases have extreme values for the waiting time comparing to the expected ones and this waiting time can even transcend the recommended waiting time value for colours that represent less urgent cases. Comparing waiting times among themselves, we can see that:

 - 60.6% (656) of the red cases waited more than the time proposed for the orange colour, 3.5% (38) for the yellow colour, 1.6% (17) for the green colour and 0.9% (10) for the blue colour;

 - 6.8% (1220) of the oranges cases waited more than it was proposed for the yellow colour, 2.1% (377) for the green and 0.8 (137)% for the blue one;

 - 13.3% (10384) of the yellow cases waited mo than the limit suggested for the green colour and 2.6% (2058) for the blue colour;

 - 3.3% (1836) of the green cases waited more than the recommended time for the blue colour.

To achieve all these results we had to select those cases that have a correct value for the variable “waiting time”.

From the 336526 individuals in the sample, 156556 (46,5%) of them have estimated values for the waiting time. However, 179970 (53,5%) of them don’t have the requested information to estimate the waiting time. As a possible explanation for this high percentage of missing cases can be the lack of registration of medical assessment. The variable “codified waiting time”, created in the database, includes the missing cases mentioned previously and also the cases which have an incorrect waiting time: negative time (errors) or superior to 24hours (possible errors). Because of that the number of missing cases of the variable “codified waiting time” (180049) is bigger than the number of missing cases of the variable “waiting time”. Therefore, only the cases included in the variable “codified waiting time” can be used to analyze if the waiting time exceeded the expected time or not. From the cases with this variable, 15 % has exceeded the expected time and 0,9% couldn’t be measured because they correspond to the cases catalogued with the white colour.

                Figure 5 - Percentage of cases without the time of the medical assessment.

By analysing the table above (table 2), we can conclude that as the urgency of the cases increase, the number of cases with the time of medical assessment also increase. Moreover as the urgency of the cases gets higher, the number of cases with a waiting time exceeded increase too.  It is important to emphasize that the blue cases (less serious) are the ones that have a high number of cases without registration of medical observation. On the other hand the more serious cases, the red ones, have a short percentage of cases which don’t have a registration of medical observation. In fact, this perspective is the opposite that was expected because if the blue and green ones can wait more time there are all the conditions to do a correct registration. The opposite situation should happen with the red and orange cases.

            As one of the outcomes of an urgency episode we have death and, pursuing on the accomplishment of our objectives, there’s the need to understand if the rate of death is superior in the patients who waited more time in the Urgency.

In 80,5% of the cases which waited more than the standard waiting time of the correspondent colour the patient ended up dying. On the other side, 67,6% of the patients that didn’t wait more than the expected, didn’t die.

            In order to study the relationship between the death rate in each color and the waiting times, it was necessary to hold two filters based on the variable "Time Exceeded". After that, two crosstabs were created to compare the outcome Death for each colour in both situations:        

From these results we can observe that there were 50037 cases in which the waiting time exceeded the time recommended by the MTS being 0.8% of them (408) deaths and there were also 103547 cases in which the waiting time did not exceeded the time recommended by the MTS being 0.1% of them (99) deaths. From these cases:

- All the red cases exceeded the recommended time from MTS, as it had been said before;

- In the orange colour, the death rate is higher in the cases which the waiting time exceeded the recommended time;

- In the yellow colour, the death rate is lower in the cases which the waiting time exceeded the recommended time;

- In the green and blue colour, the death rate is similar in the cases which the waiting time exceeded the recommended time.

             So, knowing that the patient died:

    Figure 8 – In the people who have died:  percentage of those who waited longer than expected vs. those who didn’t  wait longer than expected.

                 Our last objective consists of see if there are any differences in the results along the time.  Unlike what we would expect, the percentage of waiting time exceeded increased in the most recent trimesters comparing to the first ones. We also compare the variation in the mortality rate with the efficiency of MTS over time, as it is shown in following graphs:

Other analysis

Analyzing the death’s percentage by colour we conclude that the colour with higher mortality rate is red one (29 %). On the other hand the green and the blue colours haven’t deaths and the green and yellow isn’t significant at all.

Another interesting point is to estimate the mean of ages of patients who had death as the outcome of the urgency episode. The patients who died (mean=73 years) were older than the patients that didn’t died (mean=46 years), like we could expected. 

Analyzing the frequency of returns after 48 and 72 hours for each colour we observed that the less serious cases such as blue (10,1% -after 48h; 11,4% -after 72h) and green (7,9% - after 48h; 9,0% after 72h) have a lower rate of patients that return to US. In contrast, the red colour that had 4,6% of cases which returned after 48h and 5,3% after 72h.

We could also see that all the patients that had gone to the US after 48h, returned to the US after 72h.