Question

For geriatric patients requiring hospitalization, being able to walk without assistance (known as ambulation) is important for recovery. A team of medical researchers have developed a new ambulation program intended to increase the chances that geriatric patients achieve ambulation before being discharged from the hospital. The researchers conduct an initial test

of the program at two hospitals. At each location, some patients are randomly selected to participate in the program, while the remaining patients do not participate in the program. They are hoping for a successful test, so that they can apply for further research funding to administer the program at more hospitals.
The researchers obtain the following data about the initial test of their program:
Participate in program?

Participate in program?

Achieve ambulation before discharge?

NO

YES

No (# of patients)

5,950
9,005
Yes (# of patients)

5,050

1,095

Total (# of patients)

11,000

10,100

The researchers are extremely discouraged, as these initial results suggest the program doesn’t work. If anything, it appears the program is actually making it less likely that a patient achieves ambulation before discharge.
To better understand the results, the researchers take a closer look at the data and observe the following:

Hospital A
Participate in program?

Hospital B
Participate in program?

Achieve ambulation before discharge?

NO

YES

NO

YES

No (# of patients)

950

9,000

5,000
5
Yes (# of patients)

50

1,000

5,000

95

Total (# of patients)

1,000

10,000

10,000
100
The researchers are puzzled – now it appears the program does improve the chances of achieving ambulation before discharge.
Required: Why are the researchers coming to different interpretations based on how they analyze the data?

Answer 1

The researchers are coming to different interpretations based on how they analyze the data because they are not taking into account the different patient characteristics at each hospital.

In the first analysis, the researchers are simply looking at the overall numbers of patients who achieved ambulation before discharge, regardless of which hospital they were in. This gives the impression that the program is actually making it less likely that a patient will achieve ambulation, because the percentage of patients who achieved ambulation in the program group (49.5%) is lower than the percentage of patients who achieved ambulation in the non-program group (50.5%).

However, when the researchers take a closer look at the data and break it down by hospital, they see that the program is actually effective at increasing the chances of achieving ambulation at Hospital B. In Hospital B, the percentage of patients who achieved ambulation in the program group (95%) is much higher than the percentage of patients who achieved ambulation in the non-program group (5%).

The reason for this difference is likely due to the different patient characteristics at each hospital. Hospital A has a more severe patient population, with many patients who are in poor health and have multiple chronic conditions. These patients are more likely to have difficulty ambulating, even with the help of the program.

In contrast, Hospital B has a less severe patient population, with many patients who are in good health and have few chronic conditions. These patients are more likely to be able to ambulate, even without the help of the program.

As a result, the program is more effective at increasing the chances of achieving ambulation at Hospital B, where the patient population is less severe. This is why the researchers are coming to different interpretations based on how they analyze the data.

To get a more accurate understanding of the effectiveness of the program, the researchers need to take into account the different patient characteristics at each hospital. They can do this by conducting a more detailed analysis of the data, or by conducting a clinical trial that randomly assigns patients to the program group or the non-program group, regardless of which hospital they are in.

Here are some additional points that could be included in the explanation:

The researchers should also consider the length of time that patients spent in the program. It is possible that the program is effective at increasing the chances of ambulation, but only after a certain amount of time.

The researchers should also consider the types of interventions that were included in the program. It is possible that some interventions are more effective than others.

The researchers should also consider the quality of care that was provided to patients in both the program group and the non-program group. It is possible that the quality of care was better in the program group, which could have influenced the results.