(Chest. 2005;127:1140-1145.)
© 2005
American College of Chest Physicians
Lung Cancer Detection in Patients With Airflow Obstruction Identified in a Primary Care Outpatient Practice*
Joel J. Bechtel, MD, FCCP;
William A. Kelley, MD, FCCP;
Teresa A. Coons, PhD;
M. Gerry Klein, MD;
Daniel D. Slagel, MD and
Thomas L. Petty, MD, Master FCCP
* From Western Colorado Lung Center (Drs. Bechtel and Kelley), Grand Junction, CO; Saccomanno Research Institute (Dr. Coons), Grand Junction, CO; St. Mary’s Hospital (Dr. Klein), Grand Junction, CO; United Clinical Laboratories (Dr. Slagel), Dubuque, IA; University of CO Health Sciences Center (Dr. Petty), HealthONE Alliance, Denver, CO.
Correspondence to: Thomas L. Petty, MD, Master FCCP, 899 Logan St, Suite 203, Denver, CO 80203-3154; e-mail: tlpdoc{at}aol.com
 |
Abstract
|
|---|
Introduction: This prospective study describes a community-based lung cancer identification project focusing on high-risk patients who receive general care in a primary care outpatient practice. Within 1 calendar year, a simple questionnaire was completed in 1,296 patients > 50 years old to identify 430 patients at high risk of lung cancer (smoking, family history of aerodigestive tract cancer, or occupational exposures). Spirometric abnormalities were found in 126 of these patients.
Methods: Chest posteroanterior radiographs, thoracic CT scans, and sputum cytology were offered to subjects with airflow obstruction (n = 126). Eighty-eight patients underwent all tests. Thirty-eight patients refused or could not consent in a timely fashion.
Results: Six cancers were found in the screened group, and all were treated. Two more cancers were found in the nonscreened patients with airflow obstruction. Both were treated by surgical resection or radiation therapy. Costs per cancer found were $11,925 per patient.
Conclusions: Case finding in high-risk patients in a primary care population can be accomplished at a relatively low cost.
Key Words: airflow obstruction • community detection • costs • lung cancer
|
Introduction
|
|---|
Lung cancer continues to be the most common fatal malignancy in the United States in both men and women. The old dogma is that lung cancer detection by chest radiography and patient symptoms is ineffective, resulting in a 5-year survival rate of only 15%. Currently, there are no national or international guidelines for lung cancer screening. Studies from the Early Lung Cancer Action Project,12 Mayo,3 and Japan,4 however, show that chest helical CT scans are more sensitive than standard chest radiographs in the detection of lung nodules and cancer.1234 This report describes a project designed to identify individuals at high risk for lung cancer in a family practice clinic setting, and to detect lung cancer in this high-risk group.
|
Materials and Methods
|
|---|
Study Design
Patients who visited their primary care physicians for a variety of health-care issues but not primarily for respiratory complaints were asked to fill out a one-page questionnaire designed to identify individuals at high risk for lung cancer. Criteria for high risk of lung cancer were age 
50 years old, and one of the following: (1) current or ex-cigarette smoker, 30 pack-years; or (2) an occupational exposure to asbestos or coal dust; or (3) a family history for esophageal, lung, or laryngeal cancer (so-called aerodigestive cancer). Patients at high risk for lung cancer underwent simple spirometry. Those who were found to have airflow obstruction were offered lung cancer detection studies, which consisted of sputum cytology tests, chest radiographs, and chest CT scans. This prospective study was approved by the St. Mary’s Hospital Institutional Review Board, which approves research done within the St. Mary’s Hospital Clinical System and the entire Western Slope of Colorado.
Patient Population
Patients were enrolled from a busy 15-physician family practice group, Primary Care Partners (PCP) in Grand Junction, CO. Grand Junction is a rural town of 40,000 population with a drawing area of 250,000 persons. There were 52,336 patient visits to PCP from January 2001 through December 2001, the 1-year duration of the project. These included 14,459 visits by patients > 50 years old; 1,296 questionnaires were completed (Fig 1
). Four hundred thirty of these questionnaires were indicative of high risk. From these 430 questionnaires, 126 patients showed spirometric abnormalities; of these, 88 high-risk patients were screened (80 by smoking history, 6 by occupational history, and 2 by family history) [one lung cancer and one esophageal cancer]. Many patients were seen more than once during the year.
Testing Methods
Spirometry was performed by trained nurses in the PCP offices on the same day of the office visit. American Thoracic Society standards were followed, and airflow obstruction was diagnosed in patients with FEV1 < 70% of predicted and FEV1/FVC < 70%. Posteroanterior and lateral chest radiographs were performed in standard fashion, and overread by two physicians. Sputum cytology examinations consisted of induced and spontaneous coughed specimens prepared and analyzed using the Saccomanno technique. All sputa were interpreted by qualified sputum cytopathologists (G.K. and D.S.). Chest CT scans were performed on individuals using a helical CT scanner (GE Lightspeed QXI; GE Healthcare Technologies; Waukesha, WI), taking noncontrast, 7.5-mm, sequential images. These were independently overread by a second radiologist and a pulmonologist who did not know the results of the initial readings. Patients with suspicious sputum cytology, abnormal chest radiographic findings, or worrisome nodules on CT scan were either monitored by their family physician with follow-up studies or referred to a pulmonologist for evaluation.
|
Results
|
|---|
Eighty-eight of the 126 patients who qualified for cancer "screening" (imaging and cytology) based on high-risk history and airflow obstruction were studied. Two of the remaining 38 patients could not be scheduled for the consenting process in a timely fashion, but lung cancer was subsequently diagnosed; others refused to undergo screening tests for personal reasons. The majority of screened patients completed all tests (Table 1
).
Thirty of the 88 screened patients showed radiologic or cytologic abnormalities requiring diagnostic procedures or follow-up studies. Primary lung cancer in six of these high-risk patients was diagnosed using surgical biopsy5 or fine-needle aspiration (FNA) biopsy (Tables 2, 3
).1 CT size for the nodular lesions is listed in Table 2 (0.9 to 2.5 cm). One endobronchial lesion was 1-mm thick with extensive local endobronchial spread. Diagnoses were made in four patients at the initial testing, while diagnoses in the other two patients (patients 5 and 6, Table 2) were made when follow-up CT scans showed growth of a suspicious nodule. Biopsies were not performed initially in these two patients because of clinical concerns; one patient showed pulmonary silicosis with multiple nodules on CT scan (patient 5), and the other patient was an elderly woman with advanced emphysema (patient 6). All six patients with cancer underwent surgical resection or radiation therapy for cure. The remaining 24 of the 30 patients with screening abnormalities showed no evidence of cancer on biopsy or follow-up studies (Fig 1).
Of the 38 patients with spirometric abnormalities who were not screened, two additional cancers were found on radiograph follow-up. Both patients received surgical resection.
Cost Analysis
The cost for the initial screening of the 430 high-risk patients identified by the questionnaire was $50,055. The cost of diagnosis in six patients with lung cancer was $12,632, and the cost of follow-up studies for the 24 high-risk patients with abnormal initial screening tests was $8,861. The total cost to screen, follow-up abnormal test results, and diagnose lung cancer was $11,925 per patient detected with cancer (Table 4
). The costs recorded in this article are actual costs based on Medicare reimbursement rates for our region, and include hospital, radiology, and pathology technical fees and physician professional fees. Screening CT scans and cytology costs were negotiated at Medicare reimbursement rates.
Preliminary Follow-up
Patients 1, 2, and 3 are alive without evidence of recurrences 2 to 3 years after diagnosis. Patient 4 died of heart failure after coronary artery bypass 2 years after diagnosis. Patient 5 died of silicosis, COPD, and lung cancer. Patient 6 died of emphysema. One of the two patients who did not consent is alive after 2 years; one patient is dead.
A follow-up of all 430 high-risk patients both with airflow obstruction (n = 126) and with normal airflow (n = 304) is now underway. It will be the subject of a later report.
Comments
This was not a screening study for lung cancer. Instead, it attempted to find patients with previously unsuspected lung cancer in a general private practice setting.
The 6.8% lung cancer incidence (six cancers in 88 screened patients) found in this study is higher than reported elsewhere in a screening study.5 This is attributed to the following three factors: (1) only high-risk patients > 50 years old with extensive smoking history or an occupation or family history risk were studied. Lung cancer certainly occurs in patients with fewer risk factors; however, this study focused on very high-risk individuals, ie, a "case management" approach. (2) High-risk patients were screened only if airflow obstruction measured by spirometry was present. Tockman et al6 showed years ago that cancer was up to six times more common in smokers with airflow obstruction compared to smokers with normal lung function. (3) Detection studies included chest CT scan, chest radiography, and sputum cytology analysis. Chest CT scanning is a powerful tool to identify lung lesions as shown in the Early Lung Cancer Action Project,12 Mayo,3 and several Japanese studies45; CT scanning detected the cancer in five of the six patients in this study. Our group has previously shown7 that tiny endobronchial cancers not seen on radiography could be detected only by sputum cytology testing. The long-term survival of sputum-diagnosed lung cancer is > 50% after 5 years.8 One of the six cancers found here was identified only by cytology testing, with the screening CT scan and follow-up contrast CT scan findings both negative. (4) The 24 of 30 patients with screening abnormalities who did not show evidence of lung cancer on biopsy or follow-up represents an 80% false-positive rate. This is lower but of similar magnitude to the false-positive rate reported in other studies5 of as high as 98%.
Based on the above observations, we conclude that identifying high-risk patients by questionnaire, and testing those who show airflow obstruction with CT scan and sputum cytology improves the results in a high rate of lung cancer detection. The cost of $11,925 per patient with lung cancer detected includes all costs for screening, diagnostic testing, and follow-up of abnormal findings in those with and without cancer.
Arguments have been made that lung cancer screening is not cost-effective.9 These projections were made on the basis of many assumptions and concluded that the spending of $42,000 for quality-adjusted life-year gained is the most favorable projection. It has generally been accepted for prevention and therapeutic intervention that $50,000 per year of life saved could be accepted as cost-effective.10 Our study does not include a computation of cost-effectiveness, because the number of patients and the follow-up is now between 2 years and 3 years at this writing, ie, incomplete. Also, the number of patients is small. The costs of care of lung cancer when diagnosed by symptoms or chest radiograph observation, ie, in late stages, was nearly $50,000 per patient in 1998 dollars.11 Early diagnosis may improve on cost of later diagnosis of lung cancer.
The patients participating in the study were visiting their primary care physicians for routine health-care concerns, annual physical examinations, or a variety of acute and chronic illnesses. Only a few (estimated 10%) had respiratory complaints. The physicians, office staff, and patients were mostly enthusiastic about the study and concluded that the process involving questionnaires and spirometry performed at the time of the clinic visit was not too disruptive to patient care. No additional costs for the routine office visit and questionnaire were charged. This is a model that could be employed in many small or large family physician offices, with some adjustments.
The late outcome of all of the six patients (mean age, 72.6 years) [three patients with stage Ia lung cancer, two patients with stage IIIa lung cancer, and one patient with three cancers] will be determined by follow-up. A continuation of the long-term follow-up of the entire 430 patients at high risk is currently underway. The patients who were originally screened will undergo repeat CT and sputum cytology at year 3 and year 5 of the follow-up. The nonscreened subjects with airflow obstruction and those with risk factors but normal airflow will be followed up during their scheduled clinic visits for symptoms of lung cancer. Some will undergo chest radiographs for various indications. This follow-up is in keeping with the present standard of care in which no general screening is recommended.
The strengths and weaknesses of the study deserve comment. The main strength is that the study was done in offices of primary care practitioners during routine appointments for other medical problems (usually not respiratory symptoms). This was not a lung cancer screening study. It was a case-finding study limited to high-risk patients. A simple one-page standard questionnaire was used. By agreement, the study was limited to 1 calendar year, but the study enrolled all patients fulfilling the high-risk definition.
Weaknesses include of the study are as follows: (1) 38 of the 126 high-risk patients with airflow disorders refused or otherwise did undergo the screening tests. Two additional cancers were found in this group who did not consent in a timely fashion. All of these patients are available for follow-up, however. (2) The primary clinical diagnosis for each patient seen, which was sometimes multiple, was not systematically recorded. (3) It is possible that the Grand Junction community could be considered unique and not representative of the United States. This area is where modern sputum cytology became established, originally because of the large number of uranium workers with lung cancer in the region. Uranium has not been mined on the Colorado Plateau for nearly 30 years, but the community culture still remains alert to the possibility of diagnosing lung cancer early through sputum cytology. (4) Some costs were minimized due to agreements with St. Mary’s Hospital and Rocky Mountain Health Maintenance Organizations, but actual cost dollars were recorded. It could be argued that our costs are inaccurate, as many of the diagnostic and screening tests were negotiated to a lower fee than usually charged. It must be noted, however, that these fees are equal to Medicare-allowed fees for our area, and most physicians accept these rates. (5) Because of the practice setting study design, there is no internal comparison (control group).
In conclusion, patients at high risk for lung cancer can be identified by questionnaire and spirometry in a busy family practice clinic. Lung cancer can be detected in high-risk patients using CT scans, radiographs, and sputum cytology. Chest CT scan was the most sensitive test to identify lung cancer in our patients; standard chest radiographs added little additional information.
|
Acknowledgements
|
|---|
The authors thank Gracie Ehlert, RN, for coordinating data collection. We also wish to acknowledge the Saccomanno Research Institute, Rocky Mountain Health Plans, and Primary Care Partners of Grand Junction, CO.
|
Footnotes
|
|---|
Abbreviations: FNA = fine-needle aspiration; PCP = Primary Care Partners
Funding, in part, was provided by a grant from the Flight Attendants Medical Research Institute, Miami, FL.
Received for publication April 26, 2004.
Accepted for publication September 1, 2004.
|
References
|
|---|
- Henschke, CI, McCauley, DI, Yankelevitz, DF, et al (1999) Early Lung Cancer Action Project: overall design and findings from baseline screening. Lancet 354,99-105[CrossRef][ISI][Medline]
- Henschke, CI, Naidich, DP, Yankelevitz, DF, et al Early Lung Cancer Action Project: initial findings on repeat screenings. Cancer 2001;92,153-159[CrossRef][ISI][Medline]
- Jett, JR, Swensen, SJ, Midthun, DE, et al Screening for lung cancer in high risk groups: current status of low dose spiral CT screening and sputum markers. Semin Respir Crit Care Med 2000;21,385-392
- Koite, T, Terashima, M, Takizwa, T, et al The influence of lung cancer mass on surgical results. Lung Cancer 1999;24,75-80[CrossRef][ISI][Medline]
- Swensen, SS, Jett, JR, Sloan, JA, et al Screening for lung cancer with low-dose spiral computed tomography. Am J Respir Crit Care Med 2002;365,508-513
- Tockman, MS, Anthonisen, NR, Wright, EC, et al Airways obstruction and the risk for lung cancer. Ann Intern Med 1987;106,512-518[ISI][Medline]
- Bechtel, JJ, Kelley, WA, Petty, TL, et al Outcome of 51 patients with roentgenographically occult lung cancer detected by sputum cytologic testing: a community hospital program. Arch Intern Med 1994;154,975-980[Abstract]
- Bechtel, JJ, Petty, TL, Saccomanno, G Five year survival and later outcome of patients with x-ray occult lung cancer detected by sputum cytology. Lung Cancer 2000;30,1-7[CrossRef][ISI][Medline]
- Mahadevia, PJ, Fleisher, LA, Frick, KD, et al Lung cancer screening with helical computed tomography in older adult smokers: a decision and cost-effectiveness analysis. JAMA 2003;289,313-322[Abstract/Free Full Text]
- Tengs, TO, Adams, ME, Pliskin, JS, et al Five-hundred life-saving interventions and their cost-effectiveness. Risk Anal 1995;15,369-390[CrossRef][ISI][Medline]
- Hillner, BE, McDonald, MK, Desch, CE, et al Costs of care associated with non-small cell lung cancer in a commercially insured cohort. J Clin Oncol 1998;16,1420-1424[Abstract/Free Full Text]
TOP
Abstract
Introduction
