Randomized cancer screening trials are considered the best approach for evaluating new cancer screening technologies. Unfortunately, these trials may underestimate the efficacy of a new screening technique if study participants fail to comply with screening, diagnostic confirmation, or treatment. In this study, we propose to examine the factors associated with cancer screening nonadherence in a large, multi-center lung cancer screening trial. This study will be conducted using existing data collected as part of the American College of Radiology (ACRIN) arm of the National Lung Screening Trial (NLST). ACRIN is responsible for screening 18,843 participants at 23 sites across the United States. These participants are currently being followed to determine whether screening with low-dose spiral computed tomography (CT) or chest X-ray (CXR) reduces lung cancer mortality. We will focus on participants who receive a diagnosis at the baseline screening exam that is negative or false positive for lung cancer (meaning that the screening exam is positive, but further testing indicates that the screening abnormality is benign).
Our first aim i n this trial is to determine whether participants who receive a diagnosis that is false positive for lung cancer at the baseline screening exam are less likely to return for screening at their next scheduled screening exam than are participants who receive a screening result negative for lung cancer.
Our second aim i n this trial is to identify those participant characteristics, measured before the baseline screening exam, that are associated with nonadherence at the next scheduled screening exam. In this phase of the study, we plan to examine the influence of demographic factors, including sex, age, race, education level, income, area of the country, as well as health history, family history of lung cancer, smoking status and history, health related quality of life indices, and perceived risk of lung cancer. Obtaining information on characteristics associated with noncompliance is useful for two reasons: 1) Information on characteristics associated with noncompliance can be used to better target interventions and reduce nonadherence in future clinical trials and clinical practice, and 2) The standard approach to analyzing clinical trials is the intent to treat (ITT) analysis, in which the outcome of interest is compared between the two randomization groups regardless of participant noncompliance. Using an ITT analysis, however, assumes that the noncompliance rate observed in the trial will be equal to that in the communities in which the intervention will be implemented. This is often not the case. In screening trials, with many participants and few anticipated cancer cases, noncompliance by even a small proportion of participants with screening, follow-up, or treatment, may lead to inaccurate estimates of the efficacy of the screening exam. This study focuses on noncompliance at the screening stage. Future work will examine noncompliance with diagnostic follow-up and treatment and use the information on noncompliance to developed models that estimate the potential effectiveness of lung cancer screening under different conditions and in different target populations. This study proposes to examine factors that are associated with cancer screening nonadherence. Information on the reasons for nonadherence may be used to develop tailored interventions designed to improve adherence in future clinical trials and in clinical practice. This information may also be used to develop models that estimate the potential effectiveness of lung cancer screening under different conditions and in different target populations.
This study proposes to examine factors that are associated with cancer screening nonadherence. Information on the reasons for nonadherence may be used to develop tailored interventions designed to improve adherence in future clinical trials and in clinical practice. This information may also be used to develop models that estimate the potential effectiveness of lung cancer screening under different conditions and in different target populations.
