Screening
Conditions
Keywords
survival rates, risk communication, screening counselling, screening recommendation, understanding of medical risk
Brief summary
The probably most commonly used measure for expressing the pay-offs of early detection and treatment are survival rates. Yet, over time and groups this metric comes with several biases and thus, is not reliable for judging such benefits. Epidemiologists recommend using reduction of disease-specific mortality rates instead, which is unbiased. The purpose of the study is to investigate how primary care physicians understand and use different survival measures for determining the benefit of cancer screening tests.
Interventions
None listed
Sponsors
Dartmouth-Hitchcock Medical Center
Max Planck Institute for Human Development
Study design
Observational model
ECOLOGIC_OR_COMMUNITY
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
30 Years to 65 Years
Healthy volunteers
No
Inclusion criteria
* primary care physicians (internal, general, and family medicine physicians)
Exclusion criteria
* all other types of physicians
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Number of Physicians (=Participants) Recommending the Screening | 25 minutes (mean duration of the survey) | The aim of the study was to learn how different medical cancer screening statistics would influence doctors' recommendation behavior and their effectiveness judgments of screening tests. For that reason the online survey study presented physicians with four different medical statistics (e.g., 5-year survival) within four successive scenarios and asked after each scenario whether they would recommend the screening to a (hypothetical) patient given the data. Options to answer are: Definitely yes, Probably yes, Probably no, Definitely no, Can't decide. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Number of Physicians (= Participants) Assuming a Benefit of Screening | 25 minutes (mean duration of the survey) | Physicians are faced with four different medical statistics about the effect of screening (e.g., 5-year survival) within four successive scenarios and after each scenario asked whether they assume the screening to be beneficial given the statistical information. Options to answer are: yes, no, can't decide. If yes, then participants are further asked to describe this benefit by the following categories: Very large, large, moderate, small, very small. |
Countries
Germany
Participant flow
Recruitment details
Sample frame is the Harris Interactive Physician Panel Harris Interactive AG will draw a simple random sample U.S. internal and family medicine physicians from their Physician Panel and e-mail them an invitation and a link to the online-survey
Pre-assignment details
Inclusion criteria: Physicians in internal, family and general medicine. Exclusion criteria: All physicians other than internal, family and general medicine physicians are excluded from participation because these usually are not offering cancer screening to there patients in the setting of primary care
Participants by arm
| Arm | Count |
|---|---|
| Mortality*Incidence*5-year Survival*Early Detection Rate The study-conducted as an online survey study-investigated the influence that different medical statistics such as 5-year survival rates would have on physicians' recommendation behavior for screening and on their judgment of screening's effectiveness. The survey introduced hypothetical scenarios in which a hypothetical patient was requesting a physician's advice on whether to have a screening test. To make that decision the participants (=physicians) were presented with different statistics and then asked if they would recommend the screening to the hypothetical patient and how effective they think the screening would be in reducing cancer mortality. The online survey did not ask any sensitive data. | 778 |
| Total | 778 |
Withdrawals & dropouts
| Period | Reason | FG000 |
|---|---|---|
| Overall Study | Did not match target population | 94 |
| Overall Study | Did not respond | 65 |
| Overall Study | Hit the survey after closure | 79 |
| Overall Study | Inadvertent exclusion | 239 |
Baseline characteristics
| Characteristic | Mortality*Incidence*5-year Survival*Early Detection Rate |
|---|---|
| Age, Categorical <=18 years | 0 Participants |
| Age, Categorical >=65 years | 0 Participants |
| Age, Categorical Between 18 and 65 years | 778 Participants |
| Age Continuous | 49 years STANDARD_DEVIATION 11.6 |
| Region of Enrollment United States | 778 participants |
| Sex: Female, Male Female | 158 Participants |
| Sex: Female, Male Male | 620 Participants |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | — / — |
| other Total, other adverse events | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 |
Outcome results
Primary
Number of Physicians (=Participants) Recommending the Screening
The aim of the study was to learn how different medical cancer screening statistics would influence doctors' recommendation behavior and their effectiveness judgments of screening tests. For that reason the online survey study presented physicians with four different medical statistics (e.g., 5-year survival) within four successive scenarios and asked after each scenario whether they would recommend the screening to a (hypothetical) patient given the data. Options to answer are: Definitely yes, Probably yes, Probably no, Definitely no, Can't decide.
Time frame: 25 minutes (mean duration of the survey)
Population: We calculated that a sample size of 300 physicians was needed to have 90% power to detect differences of 20% or higher in the proportion of respondents correctly answering questions about the different cancer statistics (2-sided alpha of .05).
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Mortality*Incidence*5-year Survival*Early Detection Rate | Number of Physicians (=Participants) Recommending the Screening | 301 participants |
Comparison: Study was exploratory, thus no hypotheses had been formalized beforehand.~To analyze repeated measures outcomes (e.g., effect of the four different statistic scenarios on doctors' recommendation of screening, their judgment of screening's effectiveness, etc.), we used the McNemar chi-square test and the Wilcoxon signed-rank test.~To test for order effects (scenarios were randomly presented)Pearson's chi-square test and the Mann-Whitney U test were used.p-value: <0.05Chi-squared
Secondary
Number of Physicians (= Participants) Assuming a Benefit of Screening
Physicians are faced with four different medical statistics about the effect of screening (e.g., 5-year survival) within four successive scenarios and after each scenario asked whether they assume the screening to be beneficial given the statistical information. Options to answer are: yes, no, can't decide. If yes, then participants are further asked to describe this benefit by the following categories: Very large, large, moderate, small, very small.
Time frame: 25 minutes (mean duration of the survey)