Publication

Article

The American Journal of Managed Care
September 2023
Volume 29
Issue 9

Factors Associated With Lung Cancer Risk Factor Documentation

This cross-sectional observational study found several factors associated with whether a patient had sufficient lung cancer risk factor documentation in the electronic health record.

ABSTRACT

Objectives: To identify factors associated with the minimum necessary information to determine an individual’s eligibility for lung cancer screening (ie, sufficient risk factor documentation) and to characterize clinic-level variability in documentation.

Study Design: Cross-sectional observational study using electronic health record data from an academic health system in 2019.

Methods: We calculated the relative risk of sufficient lung cancer risk factor documentation by patient-, provider-, and system-level variables using Poisson regression models, clustering by clinic. We compared unadjusted, risk-adjusted, and reliability-adjusted proportions of patients with sufficient smoking documentation across 31 clinics using logistic regression models and 2-level hierarchical logit models to estimate reliability-adjusted proportions across clinics.

Results: Among 20,632 individuals, 60% had sufficient risk factor documentation to determine screening eligibility. Patient-level factors inversely associated with risk factor documentation included Black race (relative risk [RR], 0.70; 95% CI, 0.60-0.81), non-English preferred language (RR, 0.60; 95% CI, 0.49-0.74), Medicaid insurance (RR, 0.64; 95% CI, 0.57-0.71), and nonactivated patient portal (RR, 0.85; 95% CI, 0.80-0.90). Documentation varied across clinics. The reliability-adjusted intraclass correlation coefficient decreased from 11.0% (95% CI, 6.9%-17.1%) to 5.3% (95% CI, 3.2%-8.6%), adjusting for covariates.

Conclusions: We found a low rate of sufficient lung cancer risk factor documentation and associations of risk factor documentation based on patient-level factors such as race, insurance status, language, and patient portal activation. Risk factor documentation rates varied across clinics, and only approximately half the variation was explained by factors in our analysis.

Am J Manag Care. 2023;29(9):439-447. https://doi.org/10.37765/ajmc.2023.89354

_____

Takeaway Points

  • Lung cancer risk factor documentation is an important precursor for the determination and measurement of those eligible for lung cancer screening.
  • We found a low rate of sufficient lung cancer risk factor documentation and associations of risk factor documentation based on factors such as race, insurance status, language, and patient portal activation.
  • Risk factor documentation rates varied across clinics, and only approximately half the variation was explained by factors in our analysis.
  • Our analysis highlights potential disparities in documentation and key opportunities for improvement in clinic variability.

_____

Screening with low-dose CT (LDCT) leads to reduced lung cancer mortality among high-risk individuals.1,2 The US Preventive Services Task Force (USPSTF) first recommended lung cancer screening (LCS) in 2013 and recently expanded eligibility criteria in 2021.3,4 However, nationwide utilization of LDCT remains low, occurring in less than one-fifth of qualifying individuals.5-8 Compounding low screening rates, recent data suggest that health systems may have inadequate information to even determine screening eligibility for many patients.9

A key “upstream” and understudied aspect of LCS implementation is documentation of lung cancer risk factors within the electronic health record (EHR) to determine eligibility. Complete risk factor documentation can be defined as documenting smoking status, pack-years, and years since quit (when applicable). A seminal study revealed low and variable rates of complete risk factor documentation (44%-66%) across 4 integrated health systems.9 Risk factor documentation is often recorded in structured fields within the EHR. Unlike other recommended cancer screening, LCS requires detailed information on a behavioral risk factor. Although documenting smoking status (ie, ever/never) is straightforward and motivated by CMS EHR Incentive Programs, ascertaining the severity and duration of tobacco exposure is more challenging because health care teams must seek this information from patients and patients must know their exposure history.10 Without sufficient risk factor information, health systems are not able to identify eligible individuals for LCS. It is not known whether specific patient-, provider-, or system-level factors are associated with risk factor documentation. Better understanding of these factors can help direct health system efforts to improve risk factor documentation to identify those eligible for LCS.

The primary objective of our study is to identify factors associated with documentation of lung cancer risk factors. We used established conceptual models for LCS to prespecify factors that could affect lung cancer risk factor documentation.11-13 Additionally, we gathered input from our multidisciplinary team—consisting of a primary care physician, a pulmonologist, thoracic surgeons, population health administrators and leaders, and health services researchers—to develop an adapted conceptual model for patient-level, provider-level, and system-level factors that we hypothesized could affect lung cancer risk factor documentation (eAppendix Figure [eAppendix available at ajmc.com]).

A secondary goal of our study is to characterize the extent to which documentation of lung cancer risk factors varies among clinics in an academic health system. LCS risk factor documentation requires manually entered data in structured fields in the EHR without system prompts. Understanding clinic-level variability can help to inform the health system of uptake of standard processes for documentation and to identify high-performing clinics to glean best practices.

Finally, a third goal of our study is to determine the proportion of variability across clinics explained by measurable multilevel factors included in our analysis. We hypothesize that our measured patient-, provider-, and system-level factors account for most of the variability in lung cancer risk factor documentation.

METHODS

Study Design, Data Sources, and Participants

We conducted a cross-sectional study of individuals aged 55 to 80 years who were current or former smokers, based on structured fields within the EHR as of December 31, 2019, and empaneled to a primary care provider (PCP) within University of Washington (UW) Medicine’s primary care clinics. December 31, 2019, was defined as the index date for the study, and data were pulled on this date or during a defined look-back period for study measures. In addition to family and internal medicine, UW Medicine selectively assigns panels to other specialties such as geriatric medicine, infectious disease (including 2 HIV primary care clinics), and sports medicine. At the time of the analysis, UW Medicine primary care served approximately 330,000 individuals within western Washington, with approximately 480 PCPs working across 31 clinics. We excluded individuals with missing data on covariates of interest if it was felt that their inclusion using a missing data indicator was likely to result in statistically unreliable model estimates. We excluded those who were in a systemwide outreach program that was implemented in 2018 to improve documentation of lung cancer risk factors. The UW institutional review board approved this study (STUDY00009621). This investigation adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.14

Main Measures

Outcome. We defined sufficient risk factor documentation as having the minimum necessary information to determine an individual’s eligibility for LCS based on USPSTF 2013 criteria (applicable in 2019): aged 55 to 80 years, cigarette smokers with at least 30 pack-years of exposure, and who quit within the past 15 years if a former smoker.3 For instance, an individual with a 15-pack-year smoking history and without documentation of years since quitting is ineligible for screening. Despite incomplete documentation of a lung cancer risk factor, this individual has sufficient risk factor documentation to determine LCS eligibility with the available information. We only used risk factor information documented within structured fields in the EHR. As part of a planned sensitivity analysis, we evaluated a strict definition for complete documentation of lung cancer risk factors mandating recorded values for smoking status, pack-years, and years quit among former smokers.

Variables. Patient-level factors (ie, measured at the individual patient level) included age, sex, race, ethnicity, primary insurance, preferred language, comorbidity index, activation of an electronic patient portal, and number of clinic office visits in the past 3 years. Patient portal activation was included as a surrogate of patient engagement that has been demonstrated to affect screening outcomes in other cancer screening.15 We calculated the Charlson Comorbidity Index (CCI) score using a 3-year look-back window from the study index date (January 1, 2017, to December 31, 2019) to capture comorbidities for each patient to ensure that we included all relevant diagnoses.16 The EHR was the source of patient-level factors. We included missing data labeled as “unknown” for each variable when applicable.

Provider-level factors included provider specialty (ie, family medicine, internal medicine, and other [ie, infectious disease, geriatrics, or sports medicine]), training (ie, advanced practice provider, attending physician, or resident physician), clinical full-time equivalent (cFTE), and cFTE-adjusted panel size. We evaluated provider specialty because at the time of the analysis, internal medicine and family medicine had differing guideline recommendations for LCS, with the American Academy of Family Physicians not recommending LCS.17,18 Panel size refers to the number of individuals who are assigned to a provider for primary care. The cFTE-adjusted panel size was calculated by dividing the provider’s panel size by their cFTE. UW Medicine provided information on provider-level variables.

We evaluated 1 system-level factor, the medical assistant (MA) to PCP ratio at each clinic, based on information collected by UW Medicine. We measured this because operationally, MAs generally collect smoking status information and input it into structured fields in the EHR, and because this measure reflects the amount of support provided to PCPs.

Statistical Analysis

We reported the frequency of sufficient smoking documentation by each patient-, provider-, and system-level factor. The unadjusted and adjusted relative risks of sufficient documentation were estimated using Poisson regression models with robust error variance estimators, clustering by clinic.19,20 We chose to report the associations using relative risks because they are intuitive to understand and are often the preferred metrics for decision makers. The adjusted model included all patient-, provider-, and system-level factors.

To characterize the extent to which documentation of lung cancer risk factors varied across clinics, we computed the unadjusted and risk-adjusted proportions of patients with sufficient smoking documentation for each clinic. We also adjusted for reliability, which accounts for the proportion of observed variation in outcomes due to chance alone, using empirical Bayes techniques.21,22 In brief, we performed logistic regression models on smoking documentation to estimate the predicted probability of sufficient risk factor documentation based on demographics. We then estimated the reliability-adjusted proportions using 2-level hierarchical logit models, where we entered the patient predicted probability as the fixed effect and the clinic identifier as the random intercept. We reported the intraclass correlation coefficients (ICCs) before and after risk adjustment, which describe the proportion of total variability accounted for by between-clinic variance. Credible intervals for point estimates following reliability adjustment were estimated using empirical Bayes techniques.22

All analyses were performed using Stata version 14 (StataCorp LP). All tests were 2-sided, and P values less than .05 were considered significant. All mean values are reported with SDs.

RESULTS

We identified 64,941 patients aged 55 to 80 years who were empaneled to a provider within the health system regardless of smoking status (Figure 1). We excluded those who were not assigned to a provider panel in designated specialties (n = 19), were assigned to a provider without primary care panel data (n = 11), were assigned to a non–primary care clinic (n = 285), were of unknown sex (n = 7), and/or were included in the systemwide outreach program described earlier (n = 4237). We excluded those who did not have a recorded smoking status or history (n = 3909) and those with never smoking history (n = 35,841). We included 20,632 individuals with current or former smoking status in the analysis.

The mean age of patients in the study was 67 years; 54% were women, 77% were White, 95% were English speaking, and 65% had activated their electronic patient portal (Table 1 [part A and part B]). Compared with the US population, a greater percentage of individuals were women and White in the study population.23 Sixty percent of those with current or former smoking history had sufficient LCS risk factor documentation (Table 2).

Sufficient risk factor documentation varied based on patient-level factors (Table 3 [part A and part B]). We conducted both unadjusted and adjusted analyses to determine factors associated with sufficient risk factor documentation (Table 3). In the multivariable analyses, sufficient risk factor documentation was associated with age older than 60 years, male sex, CCI score of 2 or greater, and frequent office visits. Sufficient risk factor documentation was inversely associated with race identified as Black/African American vs White; preferred language non-English vs English; insurance recorded as Medicaid, self-pay, other, or unknown compared with Medicare; and nonactivated vs active electronic patient portal.

Sufficient risk factor documentation did not vary considerably based on most provider-level factors in the univariable or multivariable analyses, including cFTE and adjusted panel size (Table 3). Having an assigned PCP who was not in internal medicine or family medicine (ie, geriatrics, infectious disease, or sports medicine) was inversely associated with sufficient risk factor documentation.

There was an inverse association with sufficient risk factor documentation in clinics with lower MA to PCP ratios in the unadjusted analysis; however, this association was not significant in the adjusted analysis.

Clinic Variability

The unadjusted proportion of patients with sufficient risk factor documentation ranged from 21.6% to 82.7% across 31 clinics. The reliability-adjusted proportion of patients with sufficient risk factor documentation ranged from 24.2% to 83.1%, and the ICC calculation was 11.0% (95% CI, 6.9%-17.1%), reflecting total variation accounted for by between-clinic variance (Figure 2 [A]). Adding patient characteristics to the model decreased the ICC to 7.2% (95% CI, 4.4%-11.6%), meaning that patient characteristics explained approximately 34.4% of between-clinic variance. Adding provider characteristics to the model decreased the ICC to 9.2% (95% CI, 5.7%-14.5%), or explained approximately 16.6% of between-clinic variance. Adding clinic characteristics decreased the ICC to 7.7% (95% CI, 4.8%-12.3%), or explained approximately 29.7% of between-clinic variance. Adjusting for patient-, provider-, and system-level covariates and reliability, the ICC decreased to 5.3% (95% CI, 3.2%-8.6%), or explained approximately 52.1% of between-clinic variance; the proportion of patients with sufficient risk factor documentation ranged from 39.4% to 81.4% (Figure 2 [B]).

Sensitivity Analysis

In our sensitivity analysis, measuring risk factor documentation using a strict definition (ie, complete smoking status, pack-years, and years quit if former smoker), we found similar factors associated with risk factor documentation, with a few exceptions in the adjusted analyses (eAppendix Table).

DISCUSSION

Forty percent of patients who had documented smoking history did not have sufficient documentation of risk factors to determine LCS eligibility. Patient-level factors including Black race, non-English preferred language, Medicaid or no insurance, and lack of patient portal use were inversely associated with sufficient lung cancer risk factor documentation. Risk factor documentation varied across clinics, and measured patient-, provider-, and system-level factors accounted for approximately half of the clinic-level variability in documentation. Our findings have multiple implications for health systems in improving lung cancer risk factor documentation and mitigating disparities in LCS.

Novel findings from our study include that despite near universal documentation of smoking status (94%), lung cancer risk factors were not documented sufficiently in the EHR to be able to determine LCS eligibility for a large group of patients. This finding is consistent with the results of a prior study that used data from large integrated health systems and found similar rates of sufficient documentation of lung cancer risk factors.9 The consistency of findings across integrated and academic health systems in the United States suggests a pervasive problem of inadequate risk factor documentation, with potential consequences of underidentifying those who may benefit from LCS. Health systems interested in improving LCS should engage in quality improvement efforts to better understand and improve risk factor documentation.

Inverse associations with sufficient lung cancer risk factor documentation disproportionately occurred among disadvantaged segments of the population. These findings are important because they expose potential disparities in LCS that have not been described previously and may have remained undetected, as they occur upstream of LCS. Despite well-documented racial disparities in lung cancer treatment, several studies to date have not found racial differences in LCS rates.5,7,8,24-26 However, our findings demonstrate that racial disparities could exist even if there is no difference in screening rates among those with known screening eligibility; that is, patients could be inappropriately excluded from the denominator without sufficient risk factor documentation. Health systems committed to health equity must look for potential disparities in risk factor documentation to develop strategies to ameliorate disparities in LCS among marginalized populations. These findings and recommendations are directly aligned with the call from the American Thoracic Society to address disparities in LCS eligibility and access.27 Notably, patient portal activation was a modifiable patient-level factor associated with risk factor documentation that could be included in interventions addressing these disparities, although such interventions would require thoughtful implementation, given known disparities in portal use.28-30

Third, we found that risk factor documentation varied by clinic and that the factors identified in our study explained approximately half of the variability in risk factor documentation across clinics. The clinic-level variation points to health system opportunities to streamline clinic-level structures and processes for risk factor documentation. We note that prior to the study period, a subset of clinics championed quality improvement (QI) initiatives to improve lung cancer risk factor documentation independently from the centralized health system initiative described previously.31 Top-performing clinics could be identified to collect and disseminate best practices. In terms of accounting for clinic-level variation, in addition to clinic-led QI efforts, there are several variables we did not measure that may affect lung cancer risk factor documentation. For example, we were not able to assess provider knowledge or attitudes about LCS from our data, which may affect risk factor documentation.32,33 We were also not able to measure length of primary care appointments, which varies across the health system and may be meaningful, as time has been cited as a barrier to recommending LCS.34

More research will be needed to fully understand the variablesassociated with lung cancer risk factor documentation. Such research may include qualitative interviews with various stakeholders (eg, MAs, nurses, providers, administrators, patients). The inclusion of MAs in further research is especially salient, given that they are most often inputting structured data in the EHR with limited time at the start of a visit. Given the clinic-level variability we observed, further evaluation of risk factor documentation processes in high-performing clinics could also inform promising approaches and further research opportunities. Without this knowledge, it will be difficult to design highly effective interventions to equitably maximize sufficient lung cancer risk factor documentation and subsequent LCS. Expanded eligibility in the most recent USPSTF guidelines and potential disruptions in LCS risk factor documentation caused by the COVID-19 pandemic adds impetus to this work.4

Limitations

Our study has several limitations. First, we used 2013 USPSTF criteria to determine our at-risk population eligible for screening.3 Since our analysis, the USPSTF in 2021 expanded the criteria for LCS to include those who have at least a 20-pack-year smoking history within the past 15 years and to start screening at age 50 years.4 However, we do not believe the change in threshold for the severity of tobacco exposure to be eligible for screening should affect factors associated with insufficient risk factor documentation. Second, our evaluation relied on data within structured fields in the EHR. We appreciate that other methods such as natural language processing may be able to identify patients with sufficient risk factor documentation. However, clinical decision support tools and population-based outreach rely on structured data, so we think that this approach reflects best practice and identifies patients who are most likely to benefit from population-level interventions. Further, structured smoking data in the EHR have been shown in other studies to be a valid measure of smoking history.35,36 Third, our analysis likely includes unmeasured explanatory factors. For example, we did not have a measure of socioeconomic status, for which insurance status may be a surrogate marker if associated with LCS risk factor documentation. We did not include whether individuals had been seen by a pulmonologist, which could increase likelihood of LCS risk factor documentation given specialty focus and confound associations that we did identify in the multivariable analysis. We were also unable to assess provider factors that may be associated with LCS risk factor documentation (eg, provider attitudes about LCS, PCP years of experience) or clinic coordination measures (eg, team tenure) associated with improved quality.37 Finally, this analysis was exploratory in evaluating multiple potential factors with the possibility for colinearity and interactions among variables, which limits causal inference; although variables were prespecified using a conceptual model adapted from prior literature, some associations could arise from chance alone.

CONCLUSIONS

We found a low rate of sufficient lung cancer risk factor documentation in a large academic health system, and we noted associations of risk factor documentation based on patient-level factors such as race, insurance status, and language; risk factor documentation varied across clinics. Our findings have direct implications for practice improvements in LCS.

Acknowledgments

The authors thank Caroline Shevrin, MS, for assistance in editing this manuscript and James Ralston, MD, MPH, for his review of and advice for this manuscript.

Author Affiliations: Department of Medicine (LMM, MT) and Department of Surgery (DRF, DEW, DL, FF), University of Washington School of Medicine, Seattle, WA; The Comparative Health Outcomes, Policy, and Economics (CHOICE) Institute, School of Pharmacy, University of Washington (SK), Seattle, WA; UW Medicine (NA, VC), Seattle, WA; British Columbia Academic Health Science Network (DL), British Columbia, Canada; Clinical Research Division, Fred Hutchinson Cancer Research Center (MT), Seattle, WA.

Source of Funding: The work described was supported in part by Funding Opportunity Number CMS-331-44-501 from CMS and by grant number K12HS026369 from the Agency for Healthcare Research and Quality.

Author Disclosures: Dr Wood is a member of the medical advisory board for Delfi Diagnostics and has participated in a National Comprehensive Cancer Network National Lung Cancer Roundtable. Dr Triplette is employed by the Fred Hutchinson Cancer Center; has grants pending from the National Institutes of Health (National Cancer Institute) and Kuni Foundation; has received grants from the LUNGevity Foundation, American Lung Association, and National Institutes of Health; and has attended meetings of the American Thoracic Society. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (LMM, SK, DRF, VC, DEW, DCL, FF); acquisition of data (SK, DRF, NA, VC); analysis and interpretation of data (LMM, SK, DRF, VC, DEW, DCL, MT, FF); drafting of the manuscript (LMM, VC, MT, FF); critical revision of the manuscript for important intellectual content (LMM, SK, NA, VC, DEW, MT, FF); statistical analysis (SK); provision of patients or study materials (NA); obtaining funding (VC); administrative, technical, or logistic support (NA, VC, DEW, FF); and supervision (VC, DCL, FF).

Address Correspondence to: Leah M. Marcotte, MD, Department of Medicine, University of Washington School of Medicine, 4245 Roosevelt Way NE, Seattle, WA 98105. Email: leahmar@uw.edu.

REFERENCES

1. Aberle DR, Adams AM, Berg CD, et al; National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365(5):395-409. doi:10.1056/NEJMoa1102873

2. De Koning H, Van Der Aalst C, Ten Haaf K, Oudkerk M. Effects of volume CT lung cancer screening: mortality results of the NELSON randomized-controlled population based trial. J Thorac Oncol. 2018;13(suppl 10):S185. doi:10.1016/j.jtho.2018.08.012

3. Moyer VA; US Preventive Services Task Force. Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014;160(5):330-338. doi:10.7326/M13-2771

4. Krist AH, Davidson KW, Mangione CM, et al; US Preventive Services Task Force. Screening for lung cancer: US Preventive Services Task Force recommendation statement. JAMA. 2021;325(10):962-970. doi:10.1001/jama.2021.1117

5. Jemal A, Fedewa SA. Lung cancer screening with low-dose computed tomography in the United States—2010 to 2015. JAMA Oncol. 2017;3(9):1278-1281. doi:10.1001/jamaoncol.2016.6416

6. Pham D, Bhandari S, Oechsli M, Pinkston CM, Kloecker GH. Lung cancer screening rates: data from the Lung Cancer Screening Registry. J Clin Oncol. 2018;36(suppl 15):abstr 6504. doi:10.1200/JCO.2018.36.15_suppl.6504

7. Kee D, Wisnivesky J, Kale MS. Lung cancer screening uptake: analysis of BRFSS 2018. J Gen Intern Med. 2021;36(9):2897-2899. doi:10.1007/s11606-020-06236-9

8. Yong PC, Sigel K, Rehmani S, Wisnivesky J, Kale MS. Lung cancer screening uptake in the United States. Chest. 2020;157(1):236-238. doi:10.1016/j.chest.2019.08.2176

9. Gould MK, Sakoda LC, Ritzwoller DP, et al. Monitoring lung cancer screening use and outcomes at four cancer research network sites. Ann Am Thorac Soc. 2017;14(12):1827-1835. doi:10.1513/AnnalsATS.201703-237OC

10. Marcotte L, Seidman S, Trudel K, et al. Achieving meaningful use of health information technology: a guide for physicians. Arch Intern Med. 2012;172(9):731-736. doi:10.1001/archinternmed.2012.872

11. Carter-Harris L, Davis LL, Rawl SM. Lung cancer screening participation: developing a conceptual model to guide research. Res Theory Nurs Pract. 2016;30(4):333-352. doi:10.1891/1541-6577.30.4.333

12. Carter-Harris L, Gould MK. Multilevel barriers to the successful implementation of lung cancer screening: why does it have to be so hard? Ann Am Thorac Soc. 2017;14(8):1261-1265. doi:10.1513/AnnalsATS.201703-204PS

13. Beaber EF, Kim JJ, Schapira MM, et al; Population-based Research Optimizing Screening through Personalized Regimens Consortium. Unifying screening processes within the PROSPR consortium: a conceptual model for breast, cervical, and colorectal cancer screening. J Natl Cancer Inst. 2015;107(6):djv120. doi:10.1093/jnci/djv120

14. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med. 2007;4(10):e296. doi:10.1371/journal.pmed.0040296

15. Goshgarian G, Sorourdi C, May FP, et al. Effect of patient portal messaging before mailing fecal immunochemical test kit on colorectal cancer screening rates: a randomized clinical trial. JAMA Netw Open. 2022;5(2):e2146863. doi:10.1001/jamanetworkopen.2021.46863

16. Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP. The Charlson Comorbidity Index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol. 2008;61(12):1234-1240. doi:10.1016/j.jclinepi.2008.01.006

17. Mazzone PJ, Silvestri GA, Patel S, et al. Screening for lung cancer: CHEST Guideline and Expert Panel Report. Chest. 2018;153(4):954-985. doi:10.1016/j.chest.2018.01.016

18. AAFP updates recommendation on lung cancer screening. News release. American Academy of Family Physicians; April 6, 2021. Accessed May 20, 2021. https://www.aafp.org/news/health-of-the-public/20210406lungcancer.html

19. Yelland LN, Salter AB, Ryan P. Performance of the modified Poisson regression approach for estimating relative risks from clustered prospective data. Am J Epidemiol. 2011;174(8):984-992. doi:10.1093/aje/kwr183

20. Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702-706. doi:10.1093/aje/kwh090

21. Dimick JB, Ghaferi AA, Osborne NH, Ko CY, Hall BL. Reliability adjustment for reporting hospital outcomes with surgery. Ann Surg. 2012;255(4):703-707. doi:10.1097/SLA.0b013e31824b46ff

22. MacKenzie TA, Grunkemeier GL, Grunwald GK, et al. A primer on using shrinkage to compare in-hospital mortality between centers. Ann Thorac Surg. 2015;99(3):757-761. doi:10.1016/j.athoracsur.2014.11.039

23. National population by characteristics: 2010-2019. US Census Bureau. Accessed October 14, 2022. https://www.census.gov/data/tables/time-series/demo/popest/2010s-national-detail.html

24. Bach PB, Cramer LD, Warren JL, Begg CB. Racial differences in the treatment of early-stage lung cancer. N Engl J Med. 1999;341(16):1198-1205. doi:10.1056/NEJM199910143411606

25. Sineshaw HM, Wu XC, Flanders WD, Osarogiagbon RU, Jemal A. Variations in receipt of curative-intent surgery for early-stage non-small cell lung cancer (NSCLC) by state. J Thorac Oncol. 2016;11(6):880-889. doi:10.1016/j.jtho.2016.03.003

26. Corso CD, Park HS, Kim AW, Yu JB, Husain Z, Decker RH. Racial disparities in the use of SBRT for treating early-stage lung cancer. Lung Cancer. 2015;89(2):133-138. doi:10.1016/j.lungcan.2015.05.002

27. Rivera MP, Katki HA, Tanner NT, et al. Addressing disparities in lung cancer screening eligibility and healthcare access. an official American Thoracic Society statement. Am J Respir Crit Care Med. 2020;202(7):e95-e112. doi:10.1164/rccm.202008-3053ST

28. Gordon NP, Hornbrook MC. Differences in access to and preferences for using patient portals and other eHealth technologies based on race, ethnicity, and age: a database and survey study of seniors in a large health plan. J Med Internet Res. 2016;18(3):e50. doi:10.2196/jmir.5105

29. Wallace LS, Angier H, Huguet N, et al. Patterns of electronic portal use among vulnerable patients in a nationwide practice-based research network: from the OCHIN practice-based research network (PBRN). J Am Board Fam Med. 2016;29(5):592-603. doi:10.3122/jabfm.2016.05.160046

30. Veinot TC, Mitchell H, Ancker JS. Good intentions are not enough: how informatics interventions can worsen inequality. J Am Med Inform Assoc. 2018;25(8):1080-1088. doi:10.1093/jamia/ocy052

31. Peterson E, Harris K, Farjah F, Akinsoto N, Marcotte LM. Improving smoking history documentation in the electronic health record for lung cancer risk assessment and screening in primary care: a case study. Healthc (Amst). 2021;9(4):100578. doi:10.1016/j.hjdsi.2021.100578

32. Ersek JL, Eberth JM, McDonnell KK, et al. Knowledge of, attitudes toward, and use of low-dose computed tomography for lung cancer screening among family physicians. Cancer. 2016;122(15):2324-2331. doi:10.1002/cncr.29944

33. Leng J, Lei SF, Lei L, et al. Primary care providers’ knowledge, attitudes, beliefs, and practice related to lung cancer screening in five high-risk communities in New York City. J Cancer Educ. 2022;37(3):631-640. doi:10.1007/s13187-020-01857-4

34. Triplette M, Kross EK, Mann BA, et al. An assessment of primary care and pulmonary provider perspectives on lung cancer screening. Ann Am Thorac Soc. 2018;15(1):69-75. doi:10.1513/AnnalsATS.201705-392OC

35. Golden SE, Hooker ER, Shull S, et al. Validity of Veterans Health Administration structured data to determine accurate smoking status. Health Informatics J. 2020;26(3):1507-1515. doi:10.1177/1460458219882259

36. Patel N, Miller DP Jr, Snavely AC, et al. A comparison of smoking history in the electronic health record with self-report. Am J Prev Med. 2020;58(4):591-595. doi:10.1016/j.amepre.2019.10.020

37. Parkerton PH, Smith DG, Straley HL. Primary care practice coordination versus physician continuity. Fam Med. 2004;36(1):15-21.

Related Videos
dr mohamad mohty
dr ryan jacobs
Video 2 - 3 KOLs are featured in, "Assessing Limitations, Costs, Mobility and Cognition in Progressive Multiple Sclerosis"
Video 1 - 3 KOLs are featured in, "Multiple Sclerosis: Impact, Types, Pathophysiology, and Progression"
Justin Oldham, MD, MS, an expert on IPF
Justin Oldham, MD, MS, an expert on IPF
Chesahna Kindred, MD, MBA, FAAD, board-certified dermatologist, Kindred Hair & Skin Cente
A panel of experts on Alzheimer's disease
Related Content
CH LogoCenter for Biosimilars Logo