Comparing outpatient and inpatient utilization rate in physicians, informed healthcare service consumers, and non-physicians, uninformed healthcare service consumers

Abstract

utilization-altering factors were considered in the study design and analysis.The ndings validate policies to regulate service provision,plan an appropriate payment system,which is mainly fee for service payment in Iran,and increase general population-health information via trusted routes.

Background
Supplier induced demand (SID) in health economics, is the amount of demand that would have not existed if the patient was fully and symmetrically informed.[1] Due to the precise and exact de nition of McGuire 2000, "Physician-induced demand exists when the physician in uences a patient's demand for care against the physician's interpretation of the best interests of the patient".[2] The supplier, like physician, encourages the patient to demand greater quantities of health care services and shift his demand curve for the sake of supplier's own nancial and non-nancial interest.[3] Therefore, information gap between supplier and consumer of a service is the required condition that a supplier could induce the consumer a demand.In result, the informed consumer could resist the induction.This further utilization beyond the interests of the patient, results in welfare loss for the patient, which could be saved without altering or even improving health outcomes of the patient.[3] SID could be investigated using empirical evidence with different study designs like investigating the impact of physician income shocks and difference in physician fees on the amount of provided services, difference in service provision to patients with varying information, and indirect evidence like self-referral frequency in suppliers.[3] The results of demand induction differences among patients with varying information have been contentious; some reported that health care utilization was more among physicians, the informed patients who are supposed to resist the induction, contrary to the primary prediction; physicians would resist any induction and in result would utilize less amount of health services comparing to uninformed patients.[4][5][6][7][8] However, other studies reported signi cantly higher utilization rates among non-physicianpatients comparing to physician-patients.[9][10][11] Iran healthcare system consists of Primary Health Care networks as rst level, healthcare and hospital services provided by medical/health specialists as second level, and specialized, consultative healthcare for inpatient services, often through referrals from previous levels and within tertiary hospitals as third level of health service provision.The main provider payment mechanism is historically Fee for Service (FFS) and per-case payments in Iran.Like outpatient services, hospital services include common procedures reimbursed via per-case payment method and other services paid in terms of FFS method whose fee schedules are determined according to relative value units of the year.Moreover, there is no regulatory system monitoring the FFS payment which makes SID happening easier for providers as FFS and the subsequent SID has been argued to be one of main underlying reasons of the increased expenses on health in Iran.[12][13][14] A few studies have reported the existing evidences of SID in Iran healthcare market, however, none of them assessed the impact of consumer information on the amount of utilized healthcare services.[15][16][17][18][19][20] In this regard, we aimed the estimate whether informed patients differ in their amount of utilization with uninformed patients.The hypothesis is that the informed patient (physician-patient) resists the induction of a demand, thus, their utilization of healthcare services is their actual need and any further utilization in non-physician-patients may be due to induced demand, especially if the altering characteristics of actual healthcare need are considered.

Study Design:
This study was performed in the Tehran province, the capital of Iran, to compare the mean individual outpatient and inpatient utilization rates in physicians (informed consumer) and non-physicians (uninformed consumer) after controlling for conditions with probable impact on health care need and utilization.

Population:
Study population consisted of the physicians and non-physicians of the Tehran province who were randomly chosen.Subjects who did not answer the phone call and were unwilling to cooperate.In nonphysicians, subjects who had at least one physician in their close family (parents, children, siblings, spouse) or were a medical intern, were excluded as well, because their health information, and in result healthcare utilization, is probably different of other non-physicians.

Data Collection
Variables were as followings 1) Demographic: age, gender, occupation, education, marital status, living region, and income (low, lower middle, upper middle, and high based on the current currency of the country); 2) Insurance coverage: basic insurances in Iran consist of social insurances that covers more than 90% of Iranian population.
3) Access to healthcare: the least time required for achieving the nearest healthcare provision center, and number of people living permanently in the household; 4) Health status: self-perceived health status (great, good, not bad, bad) and presence of a chronic condition (including any of the following conditions for at least 3 months that is diagnosed by a physician; any metabolic, cardiovascular disease, any kind of disability, hematologic disease, psychiatric disease, chronic infection, rheumatologic disorder, or other speci ed chronic disease); We expected that age, sex, income, marital status, insurance coverage, and impaired health have a positive effect on utilization rate, opposing to minutes to the nearest healthcare center and number of people living in the household (by reducing the monetary and non-monetary sources allocated to each person in the family) which re ect an impaired access to healthcare.[5,21,22] Outpatient and inpatient utilization rates were outcomes of interest.Outpatient utilization per patient consisted of number of imaging and laboratory services, physiotherapy sessions, and medicine purchasing with or without prescription in the last month multiplied by 12 besides number of outpatient surgeries, pap smear or mammography, and endoscopies and colonoscopies in the last year for each patient.Because of physicians bring able to purchase medicines without any prescription in Iran, we included medicine purchasing without prescription as well.General physician visit, specialists visits, outpatient emergency room visits, and injections were not calculated in the nal outpatient utilization rate, because many physicians perform self-prescription [23] and including them in the nal outpatient utilization rate might have resulted in an overestimated difference.Inpatient utilization consisted of number of admissions in the last year per patient.Being admitted was de ned as an o cial inhabitancy in a hospital for at least 6 hours based on Iran Health Insurance Organization de nition.[24] For admissions, patients were asked to report how many times they have been admitted in the last year in any of the following wards; surgery, internal medicine, obstetrics and gynecology, emergency room, Intensive Care Unit or Coronary Care Unit, or any other ward.

Data Collection Tool
The questionnaire used in this study was developed for this study (supplementary le 1); The telephonesurvey questionnaire consisted of 21 questions and was designed based on the list of the services that are prone to be induced by a supplier, and therefore different in physicians and non-physicians, according to previous literature [25,26] and experts' opinion.Possible individual characteristics that could affect healthcare utilization were also included as explained in the previous section.The questionnaire was face-validated and content-validated in face to face interviews with 5 experts and 5 physicians and 5 nonphysicians with university degree.The pilot phase was performed in 21 individuals (11 physicians and 10 non-physicians) and the Cronbach's alpha based on a 10-day-apart test-retest was 0.845.[27] Additionally, the Pearson correlation coe cient was calculated for each variable; all were perfectly correlated except self-perceived health (0.79, p-value < 0.05).

Sampling Method
To our knowledge, there were no studies speci cally reporting the incidence rate ratio in physicians vs non-physicians, therefore, we were unable to calculate the sample size based on the relevant formula.Thus, we utilized the nearest sample size of the most methodologically similar study to ours.[4] Subjects were chosen by simple random selection with replacement from the relevant phone number list.Phone number lists were extracted from the Tehran province phone number list categorized by occupation.Physicians were chosen as general practitioners and specialists.Non-physicians were chosen as subjects who had any kind of job requiring university degree; including psychologists, lawyers, statisticians, accountants, etc.Further control questions examining the individual education and occupation were included in the questionnaire, but not inserted in the nal model for analysis.Physician and nonphysicians were matched on their living area based on the rst 2 digits of their phone number, due to different socioeconomical status of different living areas in Tehran.In details, for every physician who was interviewed, among the non-physicians' phone number lists, the ones who had a phone number that was the same in the rst two digit as the interviewed physician, non-physicians were contacted till they had the inclusion criteria and they were willing to cooperate.

Statistical Analysis
All baseline characteristics of the populations are reported in frequency of subgroups for categorizing variables and mean and standard deviation for numeric variables.The inpatient and outpatient utilization mean rates are reported in different subgroups of the participants as well.
For further analysis, as the outcome was count measure, we chose regression models that are speci ed for count data.In further analysis of the outpatient utilization rate, as we wanted to incorporate the overdispersion (mean and variance of the distribution were not equal), we used negative binomial model.In the analysis of inpatient utilization rate, we encountered excess zeros (the frequency of not being admitted in the last year was more than 50%) beside over-dispersion, and according to the nature of the data, we utilized hurdle negative binomial model.[28] To compare whether the outcome of hurdle regression model was different from regular negavive binomial model, we performed vuong test to the model outputs (p-value < 0.05).[29] All variables, except education, occupation, and living area because of being control variables of the inclusion criteria, were inserted in the nal model.[30] For sensitivity analysis, all analysis was done excluding outpatient surgery; because its nature is different of other outpatient services, there is a high probability that the main proportion of outpatient surgeries are cosmetic procedures, and nally, the outpatient surgeries that were utilized by physicians might have been inpatient type of surgeries but utilized as an outpatient service because physicians avoided to be admitted.Besides, the analysis were reported with exclusion of o ce visits and injections, that are probable to be performed by physicians themselves.Moreover, obstetrics and gynecology admissions were omitted from inpatient utilization, because they are mostly provided for delivery and cannot be induced or neglected.Finally, subjects who did not report any utilization, whether outpatient or inpatient, were excluded from the analysis to reduce the impact of probable underutilization in physicians on the nal comparison.Utilizing the "MatchIt" [31] and "Zelig" [32,33] packages in R for nearest neighbor matching, we also matched on the residence area before the regression results and the results did not differ signi cantly (each individual could be matched to more than one individual in the matching population).The general patterns remained the same for each step of sensitivity analysis, therefore, the details are not reported in this article.

Discussion
The hypothesis of our study was that whenever the information is symmetric between the supplier and the patient, supplier induced demand could not happen.Therefore, the rate of health care services utilized in physicians could be considered as their actual need, and any further utilization in non-physicians with similar health care needs might be caused by SID.In other words, physician-patient utilization rate is considered as "gold standard".[9] Considering physicians' utilization as gold standard does noes imply that this is the appropriate care, as the outcomes of care are not considered and evaluated, it solely indicates a standard rate for detecting evidences of possible SID.Our ndings suggested the existence of SID in Iranian healthcare market; non-physicians with asymmetric information to the provider utilized more services than physicians that were similar to the provider in terms of medical information.When adjusted for probable conditions that could alter healthcare utilization, health care utilization remined higher in non-physicians than physicians with incidence rate ratio of 1.38 (outpatient) and 3.19 (inpatient).
Previous reports yield evidence of SID in Iran healthcare market as well.The number of physicians in a particular location in Iran healthcare market was reported to have positive correlation with the amount of provided services and this correlation positively increased with higher price of the health care service both in outpatient and inpatient setting.[15] Additionally, physicians who were not a permanent salary-based employee tended to prescribe more lab tests and spend less time with their patients.
[16] When veteran patients who were and were not covered by a complementary insurance plan were compared, physicians suggested more visits to patients who had insurance coverage [17], indicating an evidence of supplierinduced demand because patients, who do not experience nancial pressure of excess utilization due to their insurance coverage, would be more probable to accept demand induction.
[18] Our results were also in accordance to previous international reports; in a population-based study of Taiwan, physicians utilized less healthcare than comparison non-physician subjects [34], female physicians and their relatives had lower cesarean delivery rate comparing to other women [35], and physicians' risk of hospitalization, all caused and major individual causes, was signi cantly lower than general population [10].In a random sample of Blue Cross/Blue Shield providers of Rhode Island, physicians had lower use of formal healthcare services comparing to chiropractors, dentists, optometrists, and podiatrists.
[36] In another report from California, non-physicians Cesarean-section rate was 7% more than physicians, controlled for demographics and clinical risks.[37] Opposing to our observations, physicians were as likely as non-physicians to consume health services in a national household survey of the Center for Health Administration Studies at the University of Chicago [5], however, they solely studied o ce visits, while other SID-susceptible outpatient services, like laboratory tests and imaging services, were not included in the analysis.Additively, the probability of using a medical care increased with patient information based on answers to a supplemental health opinion questionnaire.In spite of that, they recommended that the low consumption in poorly informed consumers might be caused by their underestimating the impact of medical services for improving their illness.
[38] Finally, physicians had higher national rates of surgical services comparing to other United States population [4].As they did not investigate the economic barriers to utilization in non-physician population, the difference might have been caused by dissimilar nancial access to services.
There are probable reasons for the observed pattern in our study; rst of all, physicians might have not utilized adequate healthcare services because physicians seem to be neglecting and reluctant in seeking appropriate medical care.[39] However, the better self-perceived health status in physicians and the much lower prevalence of any chronic disease in them suggest good health habits in our population.Secondly, the extra amount of services utilized in non-physicians might be caused by higher levels of supply induction by consumer, probably due to lower health literacy and less appropriate physician rapport.Notable, the non-physicians in our study were individuals with university degree and they possibly could communicate better than the general population with the healthcare provider.[40,41] Besides, university education is reported to positively correlate with health literacy.[42] Another possible reason contributing to more health service utilization in non-physicians might be their higher actual needs.This could be the result of better access to healthcare and bene cial health behaviors in physicians.[10] Moreover, physicians are required to be healthy for active practicing.[10] But as we conducted a multivariate analysis and adjusted the results for access to healthcare and health status, the aforementioned differences' impact on healthcare utilization is controlled and the results are adjusted for them.And nally, and according to our hypothesis, the higher utilization amount in non-physician-patients might be caused by the presence of information asymmetry and their vulnerability to be induced about the services that does not result in the best outcome with the least cost.It is mentionable that Tehran has high occupancy rate of hospital beds and the healthcare system of the city provides services to patients referred from other cities, which decreases the probability of unnecessary inpatient utilization.
First of all, the nature of the association of supplier induced demand and information asymmetry could be considered to be the same in variety of contexts, thus, we believe our results are generalizable to patients in other countries and other contexts as well and the driven conclusion is not limited to the directly studied population.In this study, we tried to meet many possible factors that could alter patients need for healthcare services including education level, age, gender, marital status, number of family members, chronic disease presence, insurance coverage,, income, healthcare access, and self-perceived health.
[26] The random sampling of individuals controlled a considerable part of selection bias.Moreover, we omitted the o ce and emergency room visits in sensitivity analysis because they are likely to be accomplished by self-prescribing in physicians [23].Furthermore, non-physicians who had a physician in their rst relatives were excluded because their utilization might be different from regular non-physicians.[5] As the survey was performed with the same procedure for all subjects, our study is free of inter-subject variations in data gathering.Yet, our study faced limitations.First, we investigated retrospective self-reported utilization cross-sectionally which might be altered by recall bias.The telephone survey, despite being easy to perform, is inclined with coverage errors, since who could not be reached by telephone, might have a lower socioeconomic status and non-responders are unknown in telephone surveys.[43] Some subpopulations are less likely to get reached by telephone, including elderly, young adults, and sick people.Telephone directories might not be comprehensive and some people may not be listed on them.Especially considering that the unlisted people might differ in the variable of interest with listed people, the result might get biased.Generally, the response rate is lower in telephone surveys which result in higher relative expense per completed interview.And nally, telephone surveys are limited in the question types, number of questions, and method of data gathering.[44] However, the key variable, utilization rate, is unlikely to be different among non-respondents and respondents, additionally, we were not able to analyze the variables affecting utilization rate due to limitations of telephone survey and they were all missing when a person was not willing to response or did not answer the phone-call.In spite of the fact that the actual amounts of utilization might differ if it was recorded objectively and via different survey methods, the method was the same in two groups and in result the difference is expected to remain constant.Moreover, comparing costs alongside with rate of utilization could have improved the strength of the conclusion.Additionally, we did not investigate mental health and dentistry services, which could have presented an opposing pattern due to high prevalence of mental disorders in physicians.[10] However, including dentistry services could have enlarged the difference because they are highly susceptible to demand induction in Iran.Available time and attitude for seeking healthcare services, other health related behaviors, service and administration access siding physical access to healthcare services could impact healthcare utilization, which were not investigated in our study.But health status and other aspects of access to healthcare was considered as their proxy.Our limited sample size and population determines the need for further and larger studies.Despite this requirement, the random chosen sample make the results almost generalizable.And nally, the cross-sectional investigation of the number of admissions per year was the only estimation of inpatient utilization that did not contain times a patient needed hospitalization, length of stay, and utilized services during hospitalization.
From a policy perspective, our ndings validate the requirement of policies addressing SID.The existence of supplier induced demand in Iranian healthcare market is mainly caused by inappropriate patients' request, excessive trust in physicians, patients' desire to utilize charge-free services more, low health literacy, and last but not least, unregulated access to general physicians and specialists.[19,20] Beside SID, not having an e cient referral system could result in non-physicians' misuse or overuse of services because it takes time and utilization for them to meet the physician with the appropriate specialty regarding their condition.Consequently, the interventions should target appropriate information increasing strategies in general population alongside with payment method reforms to change the SIDsusceptible infrastructure of Iran.Primary care coordination could decrease the inappropriate health service utilization in unaware and confused patients.[45] Furthermore, an established and reliable national resource of information and well-designed decision aids for patients could reduce the vulnerability of uninformed patients to SID. [45] Most importantly, a specialized healthcare workforce could be trained for navigating appropriate information, because patients cannot decide on their own and they might be puzzled in large amount of data provided by central resources of information.[45] Additionally, most of the information is provided by internet, which its access is lowest among lower socioeconomic groups and the elderly, who their need for information is the highest.On top of that, the provided information is probably affected by market preferences.This issue brings up the major role of government for controlling the provided information.[45] The role of the physicians must not be 7.5. Funding This study was funded by National Institute of Health Research (NIHR), Tehran University of Medical Sciences, which had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.7.6.Authors' contributions FM designed the theoretical framework of the study with the supervision of EA, RM, and IH.FM and EA designed the data collection tool and collected the data.FM cleaned the data with the supervision of EA.
FM analyzed the data with the supervision of MAM and consultancy of NA and AGh.FM framed the interpretation of the results with the supervision and guidance of RM and IH.All authors commented on the nal manuscript, FM applied the comments, and all authors approved of the version to submit.supplement4.docx

Table 2
Mean of outpatient and inpatient healthcare utilization rate in subgroups of the study population *** p-value < 0.05