Lung cancer is the most frequent cause of cancer-related mortality worldwide.1 The most prevalent type of lung cancer is non-small cell lung cancer (NSCLC), which comprises about 85% of lung cancer cases.2 As there are no cost-effective approaches to screening for lung cancer, most lung cancers are identified at an advanced stage (stage IIIB or IV).
New approaches to managing advanced lung cancer have emerged in recent years, including drugs designed to target specific genetic mutations in some tumors.3 The National Comprehensive Cancer Network (NCCN) recommends erlotinib, a receptor tyrosine kinase inhibitor of the epidermal growth factor receptor (EGFR) for first-line treatment of advanced NSCLC with EGFR mutation.4 Crizotinib is recommended to treat cancers that test positive for the anaplastic lymphoma kinase (ALK) mutation.4 Utilization of targeting agents has been found to extend the survival times for patients with the specified mutations.5 Both erlotinib and crizotinib are available at the VHA.
Previous research showed that VHA providers expressed overall favorable attitudes about genomic medicine.6 Providers perceived genomic medicine to have an important and possibly transformative role in medicine. Barriers to utilization of genomic medicine involved concerns about coordination of care, changes in workload, and increased length of patient visits. In addition to these system-level barriers, many providers had concerns about the proficiency of VHA-based practitioners to appropriately use genomic medicine.
Previous research has evaluated utilization of genomic testing and genomic-based targeted therapy (GBTT) in VA and community settings.5-8 It is unclear whether VHA-based providers are following clinical guidelines regarding genomic testing and utilization of GBTT.4 The authors set out to identify factors that impede and encourage guideline-consistent care in the management of NSCLC at the VHA. The authors specifically sought information about oncologists’ perceptions and experiences with EGFR and ALK mutation testing in patients with advanced NSCLC, as well as use of erlotinib and crizotinib in treating such patients.
Methods
This study was approved by the institutional review boards at Michael E. DeBakey VAMC in Houston, Texas and Baylor College of Medicine. In-depth qualitative interviews were conducted with VHA oncologists to examine their reported barriers and facilitators to mutation testing and prescribing of genomic-based treatment in patients with advanced NSCLC.
The sample of participants was recruited from a list of VHA medical oncologists, compiled by the study project coordinator. Investigators stratified the list by American College of Surgeons Commission on Cancer (CoC) accreditation status (yes/no) and used a stratified purposive sampling technique to recruit participants from CoC-accredited facilities and nonaccredited facilities. Recruitment and data collection occurred between March 2015 and February 2016. Oncologists were considered for inclusion if they (1) were specialists in oncology; (2) practiced at the VHA during the time of recruitment; and (3) had experience treating lung cancer at a VHA facility. During recruitment, potential participants were told that the investigators were interested in learning about oncologists’ experiences and decisions about using GBTT to treat advanced lung cancer in the VHA. Participants were scheduled for telephone-based interviews, and verbal consent was obtained prior to all interviews. Interviews ranged from 19 to 90 minutes (average, 40 min).
Recruitment was stopped at the point of thematic saturation, defined a priori as the point when 2 independent coders agreed that 3 consecutive transcripts for a given interview category (see below) rendered no new thematic concepts.9,10 Consistent with the theoretical framework developed by Cabana and colleagues, interviews were designed to elicit information about oncologists’ knowledge, attitudes, intent to use GBTT, and perceived facilitators and barriers to using GBTT in the VHA.11 Additional findings are presented elsewhere.12 The interview guide was pilot tested and revised prior to initiating data collection. All interviews were recorded, transcribed, and analyzed for content.
Analysis
Data were analyzed using framework analysis methodology, which allows for the inclusion of existing concepts as well as emergent themes within an established theoretical framework.13 Two independent coders with expertise in framework analysis independently created codes and indexed the data using Atlas.ti 6.2 (Scientific Software Development, Berlin, Germany). Disagreements about coding decisions were resolved through group consensus. Coding centered on 2 themes:
- Barriers and facilitators to mutation testing. This includes system or facility factors and testing weaknesses that act as barriers to ordering mutation testing, system or facility factors that facilitate ordering mutation testing, and oncologists’ suggestions for ways to encourage more testing in the VHA.
- Barriers and facilitators to prescribing GBTT. This includes system or facility factors that act as barriers to prescribing GBTT, system or facility factors that facilitate prescribing GBTT, and oncologists’ suggestions for ways to encourage more prescribing of GBTT in the VHA.