A Real-World Study Of CRP In Predicting The Treatment Outcome Of Antidepressants

BackgroundInflammatory processes have been connected to major depressive disorder(MDD).The CRP(C-Reactive Protein)is a widely used indicator of systemic inflammation.Many previous studies have discussed the role of the CRP in depression,which is elevated in patients with depression and can also reflect the severity of depression.Some studies have demonstrated that CRP can predict antidepressant treatment outcomes,however,they are limited to small-sample,single-drug,and strictly-restricted conditions.Therefore,we plan to see if CRP can predict antidepressant treatment outcomes in large samples in the real world.Methods1.data collection:for outpatients,patients who met the ICD-10 diagnostic criteria and were diagnosed as depressed were included,and a self-made enrollment manual was used to collect demographic information and clinical characteristics of the patients.For inpatients,we use the search function of the hospital’s electronic medical record system,selecting the psychiatric department as the inpatient department,and the discharge diagnosis with "depression" as the keyword to search and obtain information on depressed patients who had hospitalized in the psychiatric department between October 2019 to May 2021.The demographic information and clinical characteristics of the patients were also obtained from the electronic medical record records.2.CRP measurement:As a routine clinical test,CRP testing for outpatients and inpatients was done by the hospital laboratory using turbidimetry on the day of outpatient visits or during hospitalization.3.Screening and follow-up:The included patients were screened by inclusion and exclusion criteria.After screening,two trained psychiatrists followed up with the patients in an outpatient clinic or by telephone to obtain information on the type of medication,dose,duration of medication and to assess the efficacy using the Clinical Global Impressions-Improvement scale(CGI-I).4.Quality control:To reduce patient recall bias,we retrieved medication records from the hospital outpatient and inpatient electronic medical record system based on the patient’s name/outpatient number/inpatient number during the follow-up visit to confirm the type and dose of medication.Also to reduce recall bias,if patients had changed antidepressants during this period,the efficacy was assessed based on the most recent antidepressant used.5.Statistical analysis:types of medication and efficacy of 918 patients were collected,patients were divided into high CRP group(CRP ≥ lmg/L)and low CRP group(CRP<lmg/L).The drug efficacy was classified as ineffective and effective.Two kinds of comparisons were made using Kaplan-Meier survival analysis,one is comparing the difference in efficacy of different kinds of drugs in high and low CRP groups and the other is comparing the difference in efficacy of high and low CRP groups in the same kind of drugs.p<0.2 was taken as the criterion for inclusion in the multifactorial analysis,and those meeting the criteria were included in the Cox proportional risk regression model to control confounding factors.Results1.Differences in the efficacy of different types of antidepressants were compared in the high and low CRP groupsNo significant differences were found between the efficacy of SSRIs,SNRIs in the low CRP group(χ2=0.251,df=1,p=0.617>0.05,>0.2),and no significant differences were found between the efficacy of SSRIs,SNRIs in the high CRP group(x2=1.991,df=1,p=0.158>0.05,<0.2),but it met the criteria for multifactorial analysis.2.Comparing differences in efficacy between high and low CRP groups in the same kind of drugsIn patients using SSRIs,there was a significant difference in efficacy between the high-CRP and low-CRP groups(x2=4.501,df=1,p=0.034<0.05,<0.2),and meeting the criteria for multivariate analysis.In contrast,among patients on SNRIs,MT,NaSSA,there was no significant difference in efficacy between the high and low CRP groups(χ2=0.452,df=1,p=0.501>0.05>0.2),(χ2=0.340,df=1,p=0.560>0.05>0.2),(χ2=0.552,df=1,p=0.458>0.05>0.2).3.Cox proportional risk regression model to control confounding factorsThe high CRP group and SSRIs groups were entered into multifactorial analysis,and after controlling confounding factors by Cox proportional risk regression model,the results showed that SNRIs were more effective than SSRIs in the high CRP group(HR=1.652,p=0.037<0.05,95%CI:1.031-2.654);And in the SSRIs group,the low CRP group was more effective than the high CRP group(HR=1.257,p=0.047<0.05,95%CI:1.003-1.574).ConclusionIn the real world,CRP can predict the efficacy response of SSRIs to a certain extent,and patients with high CRP may have a poor response to SSRIs.Patients with high CRP may benefit more from SNRIs than those using SSRIs.