# US 2018/0322249 A1 · Treatment Recommendations Based on Drug-to-Drug Interactions > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2018/0322249 A1 - Document ID: US20180322249A1 - Status: Published application - Kind: A1 - Application number: 15/587,949 - Filed: 2017-05-05 - Published: 2018-11-08 - Earliest claimed: 2017-05-05 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Corville O. Allen; Timothy A. Bishop; Michael T. Payne; Sue S. Schmidt; Leah R. Smutzer - Web page: https://patentgrep.com/patent/US20180322249A1 ## CPC - G16H 10/60: for patient-specific data, e.g. for electronic patient records - G16H 20/10: relating to drugs or medications, e.g. for ensuring correct administration to patients - G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems - G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA ## Abstract Cognitive medical treatment recommendation mechanisms are provided. The mechanisms ingest a corpus of medical treatment content that comprises drug-to-drug interaction information. The mechanisms generate a set of drug interaction insight data structures and an initial set of candidate treatments for a medical condition of a patient based on an analysis of an electronic medical record associated with the patient. The mechanisms rank candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, where the ranking reduces rankings of candidate treatments in which drug interactions are identified in the set of drug interaction insight data structures. The mechanisms output a treatment recommendation based on the final set of candidate treatments. ## Claims ### Claim 1 A method, in a data processing system comprising at least one processor and at least one memory, the at least one memory comprising instructions executed by the at least one processor to cause the at least one processor to implement a cognitive medical treatment recommendation system, the method comprising: ingesting, by the cognitive medical treatment recommendation system, a corpus of medical treatment content, wherein the medical treatment content comprises drug-to-drug interaction information; generating, by the cognitive medical treatment recommendation system, a set of drug interaction insight data structures based on the ingested corpus of medical treatment content; generating, by the cognitive medical treatment recommendation system, an initial set of candidate treatments for a medical condition of a patient based on an analysis of one or more electronic medical records (EMRs) associated with the patient; ranking, by the cognitive medical treatment recommendation system, candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, wherein the ranking reduces rankings of candidate treatments in which drug interactions are identified in the set of drug interaction insight data structures; and outputting, by the cognitive medical treatment recommendation system, a treatment recommendation based on the final set of candidate treatments. ### Claim 2 (depends on 1) The method of claim 1, wherein generating the set of drug interaction insight data structures comprises: analyzing, by the cognitive medical treatment recommendation system, the drug-to-drug interaction information to categorize each drug-to-drug interaction referenced in the corpus of medical treatment content based on a measure of severity of the drug-to-drug interaction. ### Claim 3 (depends on 2) The method of claim 2, wherein generating the set of drug interaction insight data structures comprises: calculating, for each of the drug interaction insight data structures, an exclusion weighting factor based on a category of the drug-to-drug interaction associated with the drug interaction insight data structure. ### Claim 4 (depends on 3) The method of claim 3, wherein ranking the candidate treatments in the initial set of candidate treatments to generate a final set of candidate treatments comprises applying, to candidate treatments comprising a drug having a drug-to-drug interaction specified in a drug interaction insight data structure with another drug being taken by the patient, the exclusion weighting factor to a confidence score associated with the candidate treatment. ### Claim 5 (depends on 1) The method of claim 1, wherein the corpus of medical treatment content comprises natural language electronic documents comprising at least one of drug label documents comprising drug-to-drug interaction information, clinical statement documents comprising clinical statements from medical personnel describing drug-to-drug interactions, or patient statement documents comprising patient statements from patients describing drug-to-drug interactions. ### Claim 6 (depends on 1) The method of claim 1, wherein each drug interaction insight data structure specifies at least two drugs for which a negative result is determined to occur if the at least two drugs are administered to a patient, and an exclusion weighting value applicable to scores generated for candidate treatments that include at least one of the at least two drugs, in response to a determination that the patient is also prescribed another of the at least two drugs. ### Claim 7 (depends on 6) The method of claim 6, wherein ranking the candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, comprises, for each candidate treatment, and for each drug interaction insight data structure corresponding to a drug prescribed to the patient: determining whether the EMRs associated with the patient indicate that the patient has been prescribed another drug in the at least two drugs of the drug interaction insight data structure; and applying the exclusion weighting factor associated with the drug interaction insight data structure to a score associated with the candidate treatment in response to a determination that the EMRs indicate that the patient has been prescribed another of the at least two drugs. ### Claim 8 (depends on 1) The method of claim 1, wherein ranking the candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments comprises, for each candidate treatment in the initial set of candidate treatments: identifying at least one drug that is part of the candidate treatment; identifying, for the at least one drug, at least one drug interaction data structure associated with the at least one drug; identifying, for each drug interaction data structure in the at least one drug interaction data structure, a delta exclusion factor specified in the associated drug interaction data structure; determining, for each drug interaction data structure associated with the drug, whether a drug interaction represented by the drug interaction data structure applies to the patient based on the EMRs associated with the patient and criteria specified in the drug interaction data structure; and accumulating the delta exclusion factors for each drug interaction, represented by each of the drug interaction data structures, that apply to the patient to generate an exclusion weighting factor for the candidate treatment that is applied to a score for the candidate treatment. ### Claim 9 (depends on 8) The method of claim 8, wherein determining whether the drug interaction applies to the patient based on the EMRs associated with the patient comprises modifying the delta exclusion factor associated with a drug interaction associated with the drug, based on a matching of portions of the EMRs associated with the patient to criteria of the drug interaction specified in the drug interaction data structure. ### Claim 10 (depends on 8) The method of claim 8, wherein the delta exclusion factors specified in the drug interaction data structures are learned through automated machine learning and natural language processing of the corpus of medical treatment content. ### Claim 11 A computer program product comprising a computer readable storage medium having a computer readable program stored therein, wherein the computer readable program, when executed on a computing device, causes the computing device to implement a cognitive medical treatment recommendation system which operates to: ingest a corpus of medical treatment content, wherein the medical treatment content comprises drug-to-drug interaction information; generate a set of drug interaction insight data structures based on the ingested corpus of medical treatment content; generate an initial set of candidate treatments for a medical condition of a patient based on an analysis of one or more electronic medical records (EMRs) associated with the patient; rank candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, wherein the ranking reduces rankings of candidate treatments in which drug interactions are identified in the set of drug interaction insight data structures; and output a treatment recommendation based on the final set of candidate treatments. ### Claim 12 (depends on 11) The computer program product of claim 11, wherein the computer readable program further causes the cognitive medical treatment recommendation system to generate the set of drug interaction insight data structures at least by: analyzing, by the cognitive medical treatment recommendation system, the drug-to-drug interaction information to categorize each drug-to-drug interaction referenced in the corpus of medical treatment content based on a measure of severity of the drug-to-drug interaction. ### Claim 13 (depends on 12) The computer program product of claim 12, wherein the computer readable program further causes the cognitive medical treatment recommendation system to generate the set of drug interaction insight data structures at least by: calculating, for each of the drug interaction insight data structures, an exclusion weighting factor based on a category of the drug-to-drug interaction associated with the drug interaction insight data structure. ### Claim 14 (depends on 13) The computer program product of claim 13, wherein the computer readable program further causes the cognitive medical treatment recommendation system to rank the candidate treatments in the initial set of candidate treatments to generate a final set of candidate treatments at least by applying, to candidate treatments comprising a drug having a drug-to-drug interaction specified in a drug interaction insight data structure with another drug being taken by the patient, the exclusion weighting factor to a confidence score associated with the candidate treatment. ### Claim 15 (depends on 11) The computer program product of claim 11, wherein the corpus of medical treatment content comprises natural language electronic documents comprising at least one of drug label documents comprising drug-to-drug interaction information, clinical statement documents comprising clinical statements from medical personnel describing drug-to-drug interactions, or patient statement documents comprising patient statements from patients describing drug-to-drug interactions. ### Claim 16 (depends on 11) The computer program product of claim 11, wherein each drug interaction insight data structure specifies at least two drugs for which a negative result is determined to occur if the at least two drugs are administered to a patient, and an exclusion weighting value applicable to scores generated for candidate treatments that include at least one of the at least two drugs, in response to a determination that the patient is also prescribed another of the at least two drugs. ### Claim 17 (depends on 16) The computer program product of claim 16, wherein the computer readable program further causes the cognitive medical treatment recommendation system to rank the candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, at least by, for each candidate treatment, and for each drug interaction insight data structure corresponding to a drug prescribed to the patient: determining whether the EMRs associated with the patient indicate that the patient has been prescribed another drug in the at least two drugs of the drug interaction insight data structure; and applying the exclusion weighting factor associated with the drug interaction insight data structure to a score associated with the candidate treatment in response to a determination that the EMRs indicate that the patient has been prescribed another of the at least two drugs. ### Claim 18 (depends on 11) The computer program product of claim 11, wherein the computer readable program further causes the cognitive medical treatment recommendation system to rank the candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments at least by, for each candidate treatment in the initial set of candidate treatments: identifying at least one drug that is part of the candidate treatment; identifying, for the at least one drug, at least one drug interaction data structure associated with the at least one drug; identifying, for each drug interaction data structure in the at least one drug interaction data structure, a delta exclusion factor specified in the associated drug interaction data structure; determining, for each drug interaction data structure associated with the drug, whether a drug interaction represented by the drug interaction data structure applies to the patient based on the EMRs associated with the patient and criteria specified in the drug interaction data structure; and accumulating the delta exclusion factors for each drug interaction, represented by each of the drug interaction data structures, that apply to the patient to generate an exclusion weighting factor for the candidate treatment that is applied to a score for the candidate treatment. ### Claim 19 (depends on 18) The computer program product of claim 18, wherein the computer readable program further causes the cognitive medical treatment recommendation system to determine whether the drug interaction applies to the patient based on the EMRs associated with the patient at least by modifying the delta exclusion factor associated with a drug interaction associated with the drug, based on a matching of portions of the EMRs associated with the patient to criteria of the drug interaction specified in the drug interaction data structure. ### Claim 20 An apparatus comprising: at least one processor; and at least one memory coupled to the at least one processor, wherein the at least one memory comprises instructions which, when executed by the at least one processor, cause the at least one processor to: ingest a corpus of medical treatment content, wherein the medical treatment content comprises drug-to-drug interaction information; generate a set of drug interaction insight data structures based on the ingested corpus of medical treatment content; generate an initial set of candidate treatments for a medical condition of a patient based on an analysis of one or more electronic medical records (EMRs) associated with the patient; rank candidate treatments in the initial set of candidate treatments based on the set of drug interaction insight data structures to generate a final set of candidate treatments, wherein the ranking reduces rankings of candidate treatments in which drug interactions are identified in the set of drug interaction insight data structures; and output a treatment recommendation based on the final set of candidate treatments.