View source for Using natural language processing to identify problem usage of prescription opioids

The problem of prescription opiod abuse is well-known and wide-spread, but clinicians often make references to patient opioid abuse in unstructured free-text notes rather than structured, searchable discrete data fields. The article being reviewed here studies the use of [[Natural language processing (NLP)]] to help identify the patients whose problem might otherwise go unrecognized. <ref name="Carrell2015">Carrell DS, Cronkite D, Palmer RE, et al. Using natural language processing to identify problem usage of prescription opioids. Int J Med Inform. 2015;84(12):1057-64. http://www.ncbi.nlm.nih.gov/pubmed/?term=26456569</ref>

==Abstract<ref name="Carrell2015">Carrell DS, Cronkite D, Palmer RE, et al. Using natural language processing to identify problem usage of prescription opioids. Int J Med Inform. 2015;84(12):1057-64. http://www-ncbi-nlm-nih-gov.ezproxyhost.library.tmc.edu/pubmed/26456569</ref>==
===Background===
Accurate and scalable surveillance methods are critical to understand widespread problems associated with misuse and abuse of prescription opioids and for implementing effective prevention and control measures. Traditional diagnostic coding incompletely documents problem use. Relevant information for each patient is often obscured in vast amounts of clinical text.

===Objective===
The authors developed and evaluated a method that combines [[Natural language processing (NLP)]]  and computer-assisted manual review of clinical notes to identify evidence of problem opioid use in electronic health records ([[ EHRs | EHR]]).

===Methods===
The authors used the EHR data and text of 22,142 patients receiving chronic opioid therapy (≥70 days' supply of opioids per calendar quarter) during 2006-2012 to develop and evaluate an NLP-based surveillance method and compare it to traditional methods based on International Classification of Disease, Ninth Edition ([[ICD | ICD-9 ]]) codes. They developed a 1288-term dictionary for clinician mentions of opioid addiction, abuse, misuse or overuse, and an NLP system to identify these mentions in unstructured text. The system distinguished affirmative mentions from those that were negated or otherwise qualified. The authors applied this system to 7336,445 electronic chart notes of the 22,142 patients. Trained abstractors using a custom computer-assisted software interface manually reviewed 7751 chart notes (from 3156 patients) selected by the NLP system and classified each note as to whether or not it contained textual evidence of problem opioid use.

===Results===
Traditional diagnostic codes for problem opioid use were found for 2240 (10.1%) patients. NLP-assisted manual review identified an additional 728 (3.1%) patients with evidence of clinically diagnosed problem opioid use in clinical notes. Inter-rater reliability among pairs of abstractors reviewing notes was high, with kappa=0.86 and 97% agreement for one pair, and kappa=0.71 and 88% agreement for another pair.

===Conclusion===
Scalable, semi-automated NLP methods can efficiently and accurately identify evidence of problem opioid use in vast amounts of EHR text. Incorporating such methods into surveillance efforts may increase prevalence estimates by as much as one-third relative to traditional methods.

==Comments==
As the authors note, the use of [[Natural language processing (NLP)]] alone to identify a problem like opioid abuse is neither reasonable nor desirable. With the increasing transparency of electronic health record ([[EHR]]) documentation and the transmission of these records across a wide network of health information exchanges ([[HIE]]s), errors resulting from purely software-based identification of these and other problems may present a high risk for appropriate patient care. But a hybrid of NLP-assisted manual validation can significantly reduce the time and errors of omission of manual validation alone. The application of the hybrid method described in this article holds considerable promise for a wide-range of difficult-to-detect patient health problems.

==Related Articles==
[[Evaluating healthcare quality using natural language processing]]

[[Multi-label classification of chronically ill patients with bag of words and supervised dimensionality reduction algorithms]]

[[Use of a support vector machine for categorizing free-text notes]]

[[Visualizing unstructured patient data for assessing diagnostic and therapeutic history]]

==References==
<references/>

[[Category:Reviews]]
[[Category:HI5313-2015-FALL]]
[[Category:EHR]]
[[Category:Natural language processing (NLP)]]