This cross-sectional retrospective study utilized Natural Language Processing (NLP) to extract tobacco-use associated variables from clinical notes documented in the Electronic Health Record (EHR).
OBJECITVE:
To develop a rule-based algorithm for determining the present status of the patient’s tobacco-use.
METHODS:
Clinical notes ( 5,371 documents) from 363 patients were mined and classified by NLP software into four classes namely: “Current Smoker”, “Past Smoker”, “Nonsmoker” and “Unknown”. Two coders manually classified these documents into above mentioned classes (document-level gold standard classification (DLGSC)). A tobacco-use status was derived per patient (patient-level gold standard classification (PLGSC)), based on individual documents’ status by the same two coders. The DLGSC and PLGSC were compared to the results derived from NLP and rule-based algorithm, respectively.
RESULTS:
The initial Cohen’s kappa ( 1,000 documents) was 0.9448 (95% CI 0.9281–0.9615), indicating a strong agreement between the two raters. Subsequently, for 371 documents the Cohen’s kappa was 0.9889 (95% CI 0.979–1.000). The F-measures for the document-level classification for the four classes were 0.700, 0.753, 0.839 and 0.988 while the patient-level classifications were 0.580, 0.771, 0.730 and 0.933 respectively.
CONCLUSIONS:
NLP and the rule-based algorithm exhibited utility for deriving the present tobacco-use status of patients. Current strategies are targeting further improvement in precision to enhance translational value of the tool.
U.S. Department of Health and Human Services. U.S. Department of Health and Human Services. The Health Consequences of Smoking – 50 Years of Progress: A Report of the Surgeon General: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center. Rockville, MD: 2014.
2.
World Health Organization. WHO Report On The global tobacco epidemic, 2015 Raising taxes on tobacco. WHO Press 2015. www.who.int/tobacco (accessed November 2, 2017).
3.
WarnakulasuriyaSSutherlandGScullyC. Tobacco, oral cancer, and treatment of dependence. Oral Oncol2005; 41: 244–60. doi: 10.1016/j.oraloncology.2004.08.010.
4.
KhorRYipW-KBresselMRoseWDuchesneGForoudiF. Practical implementation of an existing smoking detection pipeline and reduced support vector machine training corpus requirements. J Am Med Inform Assoc2014; 21: 27–30. doi: 10.1136/amiajnl-2013-002090.
5.
LiuMShahAJiangMPetersonNBDaiQAldrichMC, et al. A study of transportability of an existing smoking status detection module across institutions. AMIA Annu Symp Proc2012; 2012: 577–86.
6.
HechtSS. Tobacco Smoke Carcinogens and Lung Cancer. JNCI J Natl Cancer Inst1999; 91: 1194–210. doi: 10.1093/jnci/91.14.1194.
7.
WhiteAD’AloisioANicholsHDeRooLSandlerD. Active Tobacco Smoke and Environmental Tobacco Smoke Exposure During Potential Biological Windows of Susceptibility in Relation to Breast Cancer. Cancer Epidemiol Biomarkers Prev2016; 25: 562–562. doi: 10.1158/1055-9965.EPI-16-0091.
8.
JamalADubeSRMalarcherAMShawLEngstromMC, Centers for Disease Control and Prevention (CDC). Tobacco use screening and counseling during physician office visits among adults – National Ambulatory Medical Care Survey and National Health Interview Survey, United States, 2005–2009. MMWR Suppl2012; 61: 38–45.
9.
SiuAL. Behavioral and Pharmacotherapy Interventions for Tobacco Smoking Cessation in Adults, Including Pregnant Women: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med2015; 163: 622. doi: 10.7326/M15-2023.
10.
ShimpiNBharatkumarAJethwaniMChyouP-HGlurichIBlamerJ, et al. Knowledgeability, Attitude and Behavior of Primary Care Providers Towards Oral Cancer: a Pilot Study. J Cancer Educ2016: 1–6. doi: 10.1007/s13187-016-1084-4.
11.
ClarkCGoodKJeziernyLMacphersonMWilsonBChajewskaU, et al. Identifying smokers with a medical extraction system. J Am Med Inform Assoc2008; 15: 36–9. doi: 10.1197/jamia.M2442.
12.
HazlehurstBSittigDFStevensVJSmithKSHollisJFVogtTM, et al. Natural language processing in the electronic medical record: assessing clinician adherence to tobacco treatment guidelines. Am J Prev Med2005; 29: 434–9. doi: 10.1016/j.amepre.2005.08.007.
13.
Centers for Disease Control and Prevention (CDC). Centers for Disease Control and Prevention. Meaningful use 2017. http://www.cdc.gov/ehrmeaningfuluse/introduction.html (accessed November 22, 2016).
14.
RostamiRHegdeHShimpiNPackGOlsonBAcharyaA. Oral Cancer Risk Assessment Using Machine Learning Algorithms. 2016 AADR/CADR Annu. Meet. (March 16–19, 2016) (Los Angeles), Los Angeles: Journal of Dental Research 95(A): 1486; 2016.
15.
HyunSJohnsonSBBakkenS. Exploring the ability of natural language processing to extract data from nursing narratives. Comput Inform Nurs2009; 27: 215-23-5. doi: 10.1097/NCN.0b013e3181a91b58.
16.
ValdezJRueschmanMKimMRedlineSSahooSS. An Ontology-Enabled Natural Language Processing Pipeline for Provenance Metadata Extraction from Biomedical Text (Short Paper). move to meaningful Internet Syst…: CoopIS, DOA, ODBASE Confed. Int. Conf. CoopIS, DOA, ODBASE… proceedings. OTM Confed. Int. Conf., vol. 10033, 2016, p. 699–708. doi: 10.1007/978-3-319-48472-3_43.
17.
DorrBJGaasterlandT. Exploiting aspectual features and connecting words for summarization-inspired temporal-relation extraction. Inf Process Manag2007; 43: 1681–704. doi: 10.1016/j.ipm.2007.01.008.
18.
HegdeHShimpiNPackGRostamiRAcharyaA. Smoking Status Classification Of Clinical Notes Using Natural Language Processing. 2016 AADR/CADR Annu. Meet. (March 16–19, 2016) (Los Angeles), Los Angeles: Journal of Dental Research 95(A): 0997; 2016.
19.
ZengQTGoryachevSWeissSSordoMMurphySNLazarusR. Extracting principal diagnosis, co-morbidity and smoking status for asthma research: evaluation of a natural language processing system. BMC Med Inform Decis Mak2006; 6: 30. doi: 10.1186/1472-6947-6-30.
20.
WangCAkellaR. A Hybrid Approach to Extracting Disorder Mentions from Clinical Notes. AMIA Jt Summits Transl Sci Proc AMIA Summit Transl Sci2015; 2015: 183–7.
21.
GouldMKSakodaLCRitzwollerDPSimoffMJNeslund-DudasCMKushiLH, et al. Monitoring Lung Cancer Screening Utilization and Outcomes in Four Cancer Research Network Sites. Ann Am Thorac Soc2017: AnnalsATS.201703-237OC. doi: 10.1513/AnnalsATS.201703-237OC.
22.
ReganSMeigsJBGrinspoonSKTriantVA. Determinants of Smoking and Quitting in HIV-Infected Individuals. PLoS One2016; 11: e0153103. doi: 10.1371/journal.pone.0153103.
23.
CavazzaMZweigenbaumP. Extracting implicit information from free text technical reports. Inf Process Manag1992; 28: 609–18. doi: 10.1016/0306-4573(92)90030-4.
24.
ParkAHartzlerALHuhJMcDonaldDWPrattW. Automatically Detecting Failures in Natural Language Processing Tools for Online Community Text. J Med Internet Res2015; 17: e212. doi: 10.2196/jmir.4612.
25.
VarpioLRashotteJDayKKingJKuziemskyCParushA. The EHR and Building the Patient’s Story: A Qualitative Investigation of How EHR Use Obstructs a Vital Clinical Activity. Int J Med Inform2015; 84: 1019–28. doi: 10.1016/j.ijmedinf.2015.09.004.
26.
NikfarjamAEmadzadehEGonzalezG. Towards generating a patient’s timeline: Extracting temporal relationships from clinical notes. J Biomed Inform2013; 46: S40–7. doi: 10.1016/j.jbi.2013.11.001.
27.
WangLRuanXYangPLiuH. Comparison of Three Information Sources for Smoking Information in Electronic Health Records. Cancer Inform2016; 15: 237–42. doi: 10.4137/CIN.S40604.
28.
SavovaGKMasanzJJOgrenPVZhengJSohnSKipper-SchulerKC, et al. Mayo clinical Text Analysis and Knowledge Extraction System (cTAKES): architecture, component evaluation and applications. J Am Med Informatics Assoc2010; 17: 507–13. doi: 10.1136/jamia.2009.001560.
29.
HustenCG. How should we define light or intermittent smoking? Does it matter? Nicotine Tob Res2009; 11: 111–21. doi: 10.1093/ntr/ntp010.
30.
HassmillerKMWarnerKEMendezDLevyDTRomanoE. Nondaily Smokers: Who Are They? Am J Public Health2003; 93: 1321–7. doi: 10.2105/AJPH.93.8.1321.
31.
KangNSinghBAfzalZvan MulligenEMKorsJA. Using rule-based natural language processing to improve disease normalization in biomedical text. J Am Med Inform Assoc2013; 20: 876–81. doi: 10.1136/amiajnl-2012-001173.
