A Field Experiment Study On Physicians' Priscribing Behavior In Private Clinics

BackgroundBecause of its accessibility and low management cost,private clinics are accepted by people and become an important supplement to public hospitals.They are an important part of the grass-roots medical service system in our country.However,due to the difficulty in obtaining relevant data,the issues related to the management of private clinics have been rare in the area of health management.The physician's prescribing behavior is an important part in the process of providing health services,which is directly related to the patient's health.However,due to the fact that physicians' behavior is related to the interests of clinic in current clinic service market,irrational drug prescribing and antibiotics abuse still exist.The incidence of the common cold is high,affecting people's quality of life,resulting in labor and economic losses.At present,there are many problems in the treatment of common cold in primary medical institutions.The abuse of antibiotics is one of the major problems.Although various policies and regulations and regulatory measures have been promulgated in our country to standardize and regular the practice of clinics and physicians' behavior,due to the lack of specific measures for private clinics and the implementation of related measures,the results are not satisfactory and there still are irregular medical practices in private clinics such as abuse of antibiotics and substandard behavior.Therefore,to explore exactly what factors affect the physician's prescribing behavior and what regulatory measures can effectively improve the quality of health services for private physicians,will help to formulate health policies and improve the quality of clinic health services.This article is focus on the discussion of these problems.Uncertainties in health services and information asymmetry are important features of the health services market that make the demand side of the health service-the patient and the provider-the physician establish a principal-agent relationship,but when the information is asymmetric and lacking corresponding regulatory measures,it can easily lead to the phenomenon of "moral damage" and induced demand.The behavior of physicians cannot be explained by well the "rational human" hypothesis in the traditional economics in the health service market.Therefore,the altruistic model of physicians has been proposed,but there is no empirical evidence to support it.In the health service market,the gender proportion of medical staff is changing,and the proportion of female medical workers in our country is gradually increasing.It is very necessary to explore the relationship between the physician's gender and drug prescribing behavior to make new health policies.At present,the supervision of clinics lacks corresponding measures and is difficult to implement and so on.According to Hawthorne's utility theory,when people know that they are involved in research or being observed,they will react accordingly.We propose to influence the physician's prescribing behavior by telling them we are observing them on our anonymous audit experiments and we hypothesize that this intervention will lead to changes in physician behavior and we will validate this hypothesis by using randomized controlled trials.Through the review and analysis of the literature of previous studies,we found that there still are some shortcomings in the current research on the relationship between altruism,gender,supervision intervention and physician's drug administration behavior:(1)The research on altruistic nature of physicians mainly stayed on the theoretical research level,lack of empirical support for the data.There was lack of appropriate ways to measure the physician's degree of altruism,nor to analyze the relationship between physician degree of altruism and drug prescribing behavior.(2)The combinations of physicians and patients will affect the research results,resulting in endogenous problems.The existing research on the physician's gender and drug prescribing behavior did not solve this endogeneity problem.(3)The study of interventions lacked rigorous research methods,and there were problems such as unclear interaction between interventions,the evaluation methods of intervention effect were not rigorous,and the methods of collecting data were unscientific.ObjectivesThe purpose of this study:In this study,we audited physicians' prescribing behavior anonymously,to assess the quality of their health services and to find out the problems and shortcomings through field experiments.We analyzed the relationship between altruism,gender and physicians' prescribing behavior,and further explored interventions to improve physician's prescribing behavior,and to provide scientific basis for health policymakers and decision makers making relevant health policies and measures and for improving the quality of health services in grassroots medical institutions such as private clinics.The specific goals of this study are:(1)Quantify and evaluate the diagnosis and treatment of physicians by using anonymous auditing experiment,to find out the problems existing in the process of providing medical services,and to provide a reference for health policy makers and decision makers reforming and perfect the supervision and management mechanisms of grass-roots level medical institutions.(2)To analyze the differences of physicians' behaviors among different gender physicians,and to provide empirical evidence support for verifying the gender influence on physicians.(3)Evaluating the altruistic nature of the physician through altruistic experiments,collecting the data of the physician's medical services by using the audit experiment,and then analyzing the relationship between the altruism of the physician and the physician's prescribing behavior to verify the physician-altruistic hypothesis.(4)By controlling the experimental conditions to reduce the impact of various confounding factors on the experimental results,using experimental methods to collect data,evaluating the effect of interventions by randomized controlled trials and exploring the relationship between supervising interventions and physician's medical behaviors,testing our experimental hypothesis that hawthorn effect affect physicians' behavior,to provide scientific basis for policymakers and decision makers formulating policies and measures for supervision and management of primary medical institutions.Research design and methodsWe randomly selected 96 registered clinics from the official website of Jinan Municipal Health and Family Planning Commission as sample.Through the random matching method of econometrics,the ethical problem of the protecting the private information of clinics in the experiment was solved.Experimental Design(1):Anonymous audit of physician service activities.We recruited female medical students who had certain medical knowledge as auditors and trained them to standardize auditors so that auditors state their own symptoms and communicate with physicians consistently,to avoid endogenous problems affecting the experimental results.The auditors act as a patient with common cold symptoms.Visit the clinic and recorded the physician's consultation.After leaving the clinic,fill in the self-made 'atient Assessment Questionnaire'.Experimental Design(2):To quantify the altruism of physicians through altruistic experiments.Dispatch investigators to act as common cold patients to the clinic.During the consultation,the auditors drop the wallet with contact information in advance.According to the doctor will mail back the wallet to judge the doctor's degree of altruism;doctors will be divided into altruism and the non-altruistic groups.At the same time,use the experiment(1)to audit the physician's diagnosis and treatment behavior.Finally,we analyzed the difference of physician behavior between altruistic group and non-altruistic group.In this experiment,the relationship between the physician's gender and the physician's prescribing behavior was also analyzed.Experimental Design(3):Using randomized controlled trials to verify that supervisory interventions can improve the physician's medication administration.Randomize the sample clinics into the intervention group and the control group.First conduct an anonymous audit on the two groups of clinics with reference to the experiment(1).First,an anonymous audit was conducted for the two groups of clinics with reference to experiment(1).Then sending investigators to the intervention group read out clinic intervention,tell the physicians that we will send auditors playing as patients to clinic to anonymously audit the health services quality of their clinic and leave a letter of intervention and a letter of introduction to the physician.We did not do any treatment in the control group.And then refer to the experiment(1)to conduct an anonymous audit on the two groups of clinics.Finally,we compared the diagnosis and treatment behavior change of two physicians groups.In this study,STATA12.0 software and Microsoft Office Excel 2007 software were used for data entry and statistical analysis.For continuous variables,variance analysis method was used for single factor analysis,and general linear regression analysis method was used for multiple factors analysis.We analyzed classification variables by using chi-square analysis in single factor analysis and binary logistic regression model in multiple factors analysis.Results(1)Problems in the health service of clinicsThere were irregularities in diagnosis and treatment when treating the common cold in clinics.During the consultation,only half of the physicians asked whether the patients cough(42.71%for the first time and 50.53%for the second time),and the proportion of asking whether sputum aspirated by the patients was less than one tenth(the first 5.21%,7.37%for the second);over one-third of the physicians did not measure body temperature during the examination(33.33%for the first and 37.89%for the second)and nearly one-fifth did not check the throat(17.71%for the first time and 23.16%for the second time),while less than one in ten physicians(8.33%for the first time and 8.42%for the second time)used the stethoscope to examine the patient.Only a third of the doctor asked the patient whether they have allergies before prescribing among the doctor who gave the patient medicine(39.29%39.29%for the first time,the second time),two-thirds of physicians told the patients how to use these drugs after they prescribed medicine(69.05%for the first time,77.78%for the second time),very few physician told the patient the side effects of drugs(2.38%for the first time,2.22%for the second time).In addition,about one-thirds of physicians used more than two drugs for treating the symptoms of common cold(33.66%for the first time and 29.47%for the second time).During the first audit,65.63%(63/96)of physicians prescribed antibiotics,with an average of 0.69 ± 0.53 kinds of antibiotics per physician.In the second audit,there were 65.26%(62/95)of the physicians who gave antibiotics,and each physician prescribed an average of 0.69 ± 0.55 kinds of antibiotics.In this study,the "patient"did not actually suffer from illness.The auditor's symptoms were the symptoms of the common cold and did not require antibiotics for treatment when the diagnosis was unknown.Only a certain amount of supportive treatment or symptomatic treatment can be,but over 60%of the physicians have prescribed antibiotics during the two audits and the abuse of antibiotics was very serious.In this study,we did not need to prescribe the optimal drug for the patients,but in the two audits of the clinic,the percentage of physicians prescribing drugs was over 90%(90.32%for the first time,94.74%for the second time),The physician prescribed about two kinds(the first 1.97 species,the second 2.09 species)on average,the total price was 20 yuan or more(the first visit 21.54 yuan,the second visit 20.69 yuan).The induced demand in clinic was serious.(2)Relationship between altruism and physician's prescribing behaviorThe study suggested that physicians who return wallets are more altruistic than other physicians.Only 19.79%(19/96)of clinic physicians at the sample clinic called back and mailed back to the wallet means that less than a fifth of wallets are returned.Even if the probability of other unforeseen circumstances(such as finding and returning a wallets on the spot,lack of proper time for auditors to lose wallets,and missing mailing)taking into account,there is a two-thirds ratio of physicians who did not call and mail back the purse.Therefore,in terms of the return of wallet,we think the overall altruism of clinic physicians in Jinan city is relatively low,although there is a lack of relevant comparative data.Drug prescribing behavior was related to physicians' altruism.And the more altruistic group of physicians administered the overall situation was significantly better than the non-altruistic group.The main manifestation was that the amount and price of drugs prescribed by physicians from altruistic group were significantly less than physicians from non-altruism group.The total price of prescribing medicines for altruistic group was lower than that for non-altruistic group(15.31 yuan vs 23.14 yuan).In terms of price(4.16 yuan vs 8.39 yuan),the percentage of physicians who prescribed antibiotics(47.37%vs 70.13%)and the amount of prescribing antibiotics(0.47 vs.0.74),altruistic group was lower than those for non-altruistic group.(3)Association between gender and physician's prescribing behaviorThe overall situation of female physician's medication administration was significantly better than that of male physicians,which was mainly reflected in the amount and price of prescribing medicine.Female physician prescribed medicine obviously less than the male physician.The percentage of physicians who prescribed drugs and who prescribed antibiotics was lower in female physicians that in male physicians(82.98%vs.97.83%,56.20%vs.76.60%respectively);and the total price of prescribing drugs for female physicians was lower than that of male physicians(18.56%than 24.58).(4)Intervention of physician's prescribing behaviorAccording to Hawthorne's utility theory,this study proposes to inform physicians by conducting experiments and letting physicians to know the mode of supervisory intervention and that they are being observed.We dispatched investigators to visit the clinic as an investigator of Shandong University and read out intervention letters and kept letters at the clinic.Our study found that the ratio of doctors' behavior getting better in intervention group was greater than that of the control group,embodied in the proportion of physicians who prescribing non-antibiotics(2.17%vs 19.57%),the price of non-antibiotics(32.61%versus 56.52%)and total price(32.61%vs 67.39%),the percentage of physician getting better in control group was lower than that of intervention group.Therefore,it can be concluded that the supervisory interventions proposed in this study,to a certain extent,improved the physician's prescribing behavior.Conclusion and policy recommendationsIn this study,field experiments were used to audit and assess the physicians' behavior in clinic and we also analyzed the influence of the physician's altruism,gender and supervision on the physician's drug prescribing behavior.Finally,the following conclusions were drawn:(1)There were still many problems in the treatment of common cold,such as non-standard treatment process,abuse of antibiotics,induced demand and so on.(2)Physicians with higher levels of altruism behaved better than physicians with lower level of altruism.The amount,price and proportion of altruistic group all were lower than that of non-altruistic group.(3)Female physicians behaved better than male physicians.The amount,price and proportion of female group all were lower than that of male group.(4)The prescription behavior of physicians from private clinics got better because of the Hawthorne effect.Supervision and intervention in the form of anonymous audit experiments,carried out in the name of non-governmental agencies such as university,can improve the prescribing of doctors to a certain extent.Based on the conclusion of this study,we propose the following policy recommendations:(1)Strengthen the supervision of private clinics' behaviors,and promote the application of anonymous audit interventions.(2)To improve the proportion of women's medical workers in the medical institutions at the grass-roots level.(3)To improve the level of altruism in clinics in order to improve the health benefits of the patients.(4)To strengthen the publicity,education and training of the correct treatment of common cold,and to reduce the occurrence of the phenomenon of antibiotic abuse.