The First Stages of Liberalization of Public Hospitals in Iran: Establishment of Autonomous Hospitals and the Barriers.

BACKGROUND: Liberalization and decentralization of public sector has been triggered in some developing countries and in Iran by the Ministry of Health and Medical Education (MOHME) that granted autonomy to 54 public hospitals. However, establishment of such a complex organizational reform was rather unsuccessful. We aimed to explore the obstacles and barriers caused such a failure and their mechanisms.
METHODS: Using a qualitative approach in 2013, we consulted key informants at the autonomous hospitals and their affiliating universities. Data collection was done within two phases: (i) 276 unstructured questionnaires asking respondents of barriers, and (ii) 23 semi-structured interviews from the first phase's key respondents. The first phase data were analyzed using thematic analysis and the second's by framework approach based on the frame shaped at the first phase.
RESULTS: Nine obstacles were recognized including "autonomous hospitals' board composition", "delay in announcing autonomous hospitals' charges by the MOHME", "lack of financing by the committed organizations", "poor follow up for implementation of the reform", "irregular board meetings", "lack of an external overseer", "shortage of full-time physicians", "lack of management stability", and "health insurance organizations' delayed payments".
CONCLUSION: The MOHME and insurance organizations did not pay the reform expenses. There were some competing motives as well to slow the reform or to shut it down. The stages of policy formulation and implementation were done separately in Iran, so this big organizational reform encountered serious obstacles.

Introduction

Privatization of public services has been mentioned as a way to solve public sector’s traditional problems such as inefficiency and lack of motives among its staff (1–3). Hence, preference of private over public units has been raised by many theories such as agency, property-right as well as public choice (4–7), although some have argued that public sector has been attacked by false assumptions (8). Some governments have used private sector’s mechanisms in their public sector entities under the “new public management” (9). Health sector, as one of whom traditionally managed by governments, has also been subject to some extent of privatization as well. Even nations such as the UK with longest running public health systems has shown some interests in diluting the role of the central government at their health services especially across hospitals (10). Privatization or liberalization has resulted in improved efficiency in some nations such as the USA (11) and Germany (12).

Although privatization may resolve inefficiency problem in the public sector, evidence suggest that sudden and mass liberalization of public sector could lead to dysfunctional consequences, as caused a massive fiscal shock across the former Soviet nations after the dissolution (13). Moreover, some governments believe that privatization of health sector would mean as shirking their duty of saving and improving public health stated and supported through 1978 Alma-Ata and 2004 Mumbai Declarations (14). Based on such evidence and believes some models especially one developed by Harding and Preker (15), suggest that liberalization of hospitals could be considered as a spectrum, with some intermediate steps (Fig. 1). Based on this model, at the first stage of privatization public facilities would be acknowledged as autonomous units, distinguished from the budgetary units by having greater freedom at financial management, recruitment and promotion of staff.

Fig. 1:

Harding and Preker’s (15) model of organizational reform towards corporatization

In Iran, the economy has been mainly owned and administered by the public sector. The privatization efforts were taken more seriously after 2003, when the government was allowed to privatize or decentralize of 80 percent of the state assets according to the Article 44 of the Iranian Constitution. Based on such a general policy, the Ministry of Health and Medical Education (MOHME) gradually started moving toward liberalization of public university owned hospitals in 2006 (encompassing about two-third of all 900 hospitals in Iran), with granting autonomy to 18 hospitals. Then MOHME asked all medical universities - which act on behalf of the MOHME in each province and are responsible for delivering health care, academic education at medical sciences and monitoring public and private healthcare organizations - to announce at least one public hospital from their catchment area as autonomous, delegating them the appropriate management and autonomy in 2009. MOHME stated the goals of the reform as (i) continuous quality improvement; (ii) productivity improvement; (iii) acceleration of health services delivery and (iv) increase of patient satisfaction with services across hospitals (16).

In parallel, the MOHME added another incentive for hospitals switched into autonomous with increasing their annual budget compared to the regular public hospitals. This extra budget was committed by the MOHME and two main basic insurance organizations in Iran, Social Security Insurance Organization (SSIO) and Medical Services Insurance Organization (MSIO). The insurance organizations committed to pay a double bill to autonomous hospitals. In addition, the MOHME committed to increase the budget from 1.2 times of the bill to 1.6. Hence, altogether the budget would increase by about 64 percent for converted autonomous hospitals while patients would pay only as equal as regular public hospital charges.

In response to MOHME’s order, the Iranian medical universities all over the country announced 36 teaching hospitals to undergo the reform and convert into the “board of trustees-operated” or autonomous entities, which increased the total number of the autonomous hospitals to 54. Granting autonomy to hospitals would give a wider range of freedom to the hospitals. Table 1 compares a regular public hospital with an autonomous one based on the autonomy and responsibilities delegated by the government to the latter.

Table 1:

Comparison of autonomous and regular public hospitals in terms of autonomy the Iranian government granted them

	Regular university hospitals	Autonomous hospitals
Management body	Hospital head (limited authority)	Board of trustees
Budget	The insurance organizations pay hospitals for their bills (covered services) and the MOHME pays about 1.2 times of the bills (for the salary and equipment costs).	The insurance organizations committed to pay two times as the hospitals’ bills and the MOHME 1.6 times (as an incentive).
Direct revenue	From co-insurance paid by patients and reimbursement from insurance organizations. Hospitals paid 5% of their direct revenue to the affiliating medical university for overhead costs.	From co-insurance paid by patients and a double-reimbursement from insurance organizations. Hospitals do not pay affiliating medical universities for the overhead costs.
Physicians' contract with hospitals	Physicians can work simultaneously at public and private sector.	Physicians are full-time (All part time contracts should switch into full-time within 4 years).
Payment to physicians	Fee for service + salary	Fee for service (2 times as regular hospital physicians') + salary
Hiring of staff and physicians	Only with the university's agreement	With the board of trustee's agreement
Buying or selling equipment/services	Only with the university's agreement	With the board of trustee's agreement

The board of trustees includes the chancellor of the affiliating medical university (as the head of the board), hospital head, an expert in management recommended by the hospital head, a representative of charitable people recommended by the province’s charity society, two consultants from the hospital recommended by the hospital head, a representative recommended by the provincial governor or mayor, and two members from the basic insurance organizations. These members are assigned by the chancellor for a two-year period and can be selected again. The board meetings should be held at least every 3 months with two-thirds of the members. All approved regulations by the board should be confirmed by the chancellor to be executed.

Although granting autonomy to the hospitals was started by the formal announcement of 54 hospitals as autonomous, the implementation of the reform encountered some obstacles. As a result, the reform did not spread to the remaining 500 university hospitals. Some news warned the failure of the reform (17) and even suggested cessation of the reform due to its poor implementation and misuse of regulations (18). Very small numbers of studies have conducted on autonomous hospitals reform in Iran, among which one concluded that the necessary autonomy was not transferred to the hospitals, especially at the areas of strategic management, human and physical resources management, and governance arrangement and accountability (19). Nevertheless, no study has focused on exploring the challenges the reform encountered.

Hence, this study aimed to explore the obstacles of establishing autonomous hospitals in the Iranian public health sector and to figure out how the obstacles hindered the reform.

Materials & Methods

We used a qualitative approach through 2013 for our study with two phases. At the first phase, we posted a questionnaire with open-ended questions to all medical universities and all 54-university hospitals that had been granted autonomy in Iran. The questionnaire was developed by the authors and its validity checked through content analysis by a group of experts. The questionnaire included three questions: (i) if the reform had been implemented well; (ii) what the obstacles were (if any); and (iii) how the obstacles affected the reform. The respondents were hospital heads and hospital managers (members of the board of trustees), and Quality Improvement officer at each hospital, plus one staff from each medical university in charge of administrative issues of the reform and some other key informants at the universities. Consequently, 276 paper based questionnaires were sent with a stamped empty envelope of which 202 were returned after two rounds of telephone follow up (response rate = 73%). The completed questionnaires were assessed primarily for the development of themes through thematic analysis (20).

At the second phase of data collection, we contacted selected respondents from the first phase for telephone interview. The selection of interviewees was based on the nine themes developed through the first phase. For each general theme, we contacted at least three respondents who had given answers that are more comprehensive in order to obtain in-depth information. Therefore 23 telephone interviews were undertaken (some people were interviewed for more than one theme) by one of the authors (NM). Most interviewees discussed other themes beside their own theme as well because the themes were generally connected. The average length of interviews was 29 minutes. The foci of the interview questions were: (i) What the exact obstacle(s) were for implementation of the reform at the levels of MOHME, university or hospital; (ii) When, how, and why they occurred; (iii) What the impact was. The questions went on detail by using appropriate probes. All interviews were recorded after the interviewees’ agreement and a verbal consent was obtained and recorded at the start of each interview. All recorded interviews were transcribed verbatim and coded. The general frame of themes was developed at the first phase, so we used framework method (20) for analysis of the transcripts. Nevertheless, some themes were amended. At the final stage, we contacted some interviewees, as member checking (21), in order to make sure that the developed meanings and findings were consistent with the interviewees’ original meanings.

All organizations and interviewees’ name were kept private at all stages of the study and only the researchers could access to them. The names of medical universities and interviewees were switched to codes to avoid any accidental release of identities through the study. Hence, we did not report any organization’s name in this paper.

Results

We recognized nine themes summarizing the barriers and challenges of the reform. The themes or barriers were not at a same level and some might be at a higher degree of importance and trigger other barriers, but for greater clarity, we did not merge them and reported all at a same level. We used Walt and Gilson’s (22) framework for analysis of health sector policies which incorporates our developed themes for structuring our findings. Therefore the themes, which counts as barriers, are reported under three headings of policy content, process (generally policy implementation), and context. Actors or stakeholders were not reported under a separate heading but discussed through the three-abovementioned headings. Nevertheless, we could not avoid overlaps and some themes may seem appropriate to be listed under other headings as well.

Policy Content

Many barriers that challenged the reform were due to the regulations set at the policy content. The main problem mentioned by some respondents was related to the members of the board of trustees. First, the board would be headed by the affiliating medical university’s chancellor. This meant that the affiliating medical university would have the ultimate power at the board, so the hospital would lack real autonomy. Indeed the autonomous hospitals would probably be run through the university instead of making their policies inside the hospital.

“The chancellor is head of the hospital board. This decreases motivation among other board members, as the university is still the boss. What is our autonomy then?”

Another concern was MOHME’s negligence in appointing some key members at the board. The structure lacked an expert in budgeting. As hospitals would have greater degree of financial autonomy, the board would need an expert in budgeting to handle the financial issue of the hospital. Another missing member at the board mentioned by some respondents was a nursing manager. Respondents believed that the board had certain members to support and lead physicians and administrative staff at hospital, but no one for nurses.

“The board of trustees needs some more members. I know that financial issues would need someone expert at budgeting issues. In addition, nurses would need a member at the board. They are a great part of hospitals but neglected at the current structure.”

Process (Policy implementation)

Unclearness about the charges and physicians’ share from them (fee for service) appeared as a serious problem at the first stages of implementation of the reform. In Iran, a list of charges for services is announced annually by the MOHME at the start of each year, but the MOHME did not announce any special list for the autonomous hospitals, but only for the regular public ones. This caused some problems in the autonomous hospitals in paying physicians’ share and in contracting the insurance organizations, which were supposed to reimburse the autonomous hospitals.

“We contacted the MOHME for the list of charges but they have not developed that. We have to charge like a regular hospital and now are in problem with our physicians’ fee for service.”

As explained in the introduction, the autonomous hospitals were supposed to be benefited from higher budgets, as an incentive to work autonomously. Nevertheless, they were paid only as the regular hospitals. The main insurance organizations did not pay their appointed share (2 times as they paid to the regular public hospitals). Their payment to the autonomous hospitals was delayed and then just as same they paid the regular public hospitals. Moreover, the MOHME also could not pay its own share - 1.6 times of the bill - to the hospitals. These caused hospitals could not pay fee for to their physicians’ services and ask their affiliating medical universities to assist on the delayed payments, which challenged the medical universities.

“Our autonomous hospital was not paid as was supposed to. The insurance organizations are not committed for paying their share. They paid only one times as the bill, not two times. The MOHME also could not pay its share. What a disastrous situation for the medical universities!”

The reform lacked a leading body at the MOHME to follow the execution. Most medical universities were not really forced for the implementation of the reform.

They just asked for announcing a hospital as autonomous and then no follow up was made by the MOHME. Hence, if the medical universities did not wish to establish the reform at the announced hospital, no obligation forced them to proceed.

“The ministry did not oblige us to really convey autonomy to the hospitals. We did not feel any force from their side, so everything we did was almost on the paper.”

Indeed some key informant even mentioned that the university chancellors typically did not wish the reform proceeds because granting autonomy to hospitals could decrease the chancellors’ autonomy.

“University chancellors say if the reform proceeds and all their hospitals get autonomy then what they would have to do!”

This barrier may have been resulted partly from the abovementioned factor of lack of MOHME’s follow up. Hospitals’ board of trustees should meet every three months with at least two-third of their members. However, this was not the case at most hospitals and the meetings were held less frequently and with fewer members. In some provinces, due to the members’ absence and irregular meetings, the affiliating medical university had to hold joint meetings for all its autonomous hospitals, which questions hospitals autonomy. Obviously, certain members, especially those beyond the hospitals, did not have enough motives to attend the meetings and get involved at hospitals’ policymaking.

“Our meetings are held irregularly. Usually no one attends on behalf of the mayor or commander. They have no motives. So our medical university holds joint meetings with our provinces’ three autonomous hospitals.”

Lack of an external overseer; uncertainty to proceed on the reform

Some interviewees believed that letting hospitals have more autonomy would dilute supervision on autonomous hospitals. Public hospitals are under strict supervision of the legal auditors, especially in terms of their financial issues, but for autonomous hospitals, no external overseer was appointed. This caused some hospitals make illegal decisions, so medical university chancellors get cautious about the reform implementation that slowed the reform down even more.

“Lack of any governmental supervision on [autonomous hospital] boards’ decisions is obvious. They have made, in some cases, violation of law. Sometimes they convince the chancellor for very dramatic changes. Some for example have decided to change their wards to private wards, with no insurance coverage, which ultimately limits patients’ access. Therefore, the chancellor that would take all responsibility should be very cautious on carrying out the reform issues. They may get reluctant”

Shortage of full-time physicians at public hospitals

Context (Iranian health Sector’ inherent problems)

The autonomous hospitals were supposed to change all their physicians to the full-time state, by changing their contracts or employing new physicians as full-time. However, this proved to be impossible for hospitals in practice. First, most physicians who worked at public hospitals, especially in big cities, were simultaneously working at private sector as well, so did not like to change to full-time at public hospitals. This was caused by lack of any serious rule to prohibit physicians from dual practice and the huge difference between private and public hospitals’ charges that would make private hospitals payment to their physicians much more than that of the public hospitals. Second, hospitals could not cease their contract with the employed physicians, because (i) they were not allowed legally, and (ii) they could not replace physicians with new ones, due to the lack of physicians who would be happy to work full-time. Moreover, employing physicians at certain specialties was not possible due to shortage of physicians in some specialties at the country. Therefore, the MOHME’s goal of changing 100 percent of hospitals’ physicians to full-time ones at autonomous hospitals by 2013 would be practically impossible.

“Who is willing to work full-time at public hospitals with such a low and delayed payments. Most physicians like to have their private sector activities and work only as part-time physicians at autonomous hospitals. The MOHME cannot do anything against them.”

Public hospitals usually are very prone to changes at their top levels of managerial positions and this causes difficulties in the implementation of the reform. The changes happen usually after elections or after any major change at the MOHME level.

“In the last 4 years our hospital has had 4 different heads and 3 managers. This causes establishment of the reform impossible. The managers are not confident enough to undergo the reform, because they do not know whether they would stay at their position.”

Health insurance organizations are traditionally very late in paying hospital bills. Two main health insurance organizations usually have a delay of at least six months, which makes autonomous hospitals more vulnerable (23). Most interviewees believed that with such a delay the autonomous hospitals could not afford their heavy human resources payments and so would not survive.

“Although all public hospitals are paid late by the insurance organizations, autonomous hospitals are more vulnerable. They should pay higher for their physicians.”

Discussion

In this qualitative study, we explored the obstacles that the Iranian health system encountered through liberalization of the public university owned hospitals. The findings showed different obstacles, categorized in three general headings of policy content, policymaking process (policy implementation) and the context.

As a serious obstacle categorized under policy content, “composition of the autonomous hospitals’ board”, was the main concern of most respondents. Role of the university chancellor as the head of the board would mean that the autonomous hospitals could have no real freedom and their policies would be imposed by the affiliating university and the MOHME ultimately. This shows the central state’s concerns over the first stages of liberalization and its cautious policy in delegating responsibilities to the local health sectors. Similar pattern has been seen even among developed nations such as the UK where Foundation Trusts, as autonomous entities, were indeed controlled by Primary Care Trusts that could dictate the central government’s priorities (24). Nevertheless, the whole reform was not exhausted in the UK (25) and most developed nations such as Germany (12).

Although one may argue that the chancellor is only one of the members and he/she would not necessarily impose the university’s policies on the hospital board, sensitivity to the fact that the chancellor is the head and has the ultimate power would decrease inspiration among hospital-based members. Moreover, chancellors’ concern about decrease of their authority and power probably has slowed down the process of accomplishment of the reform. Furthermore, the university would not charge a 5% overhead cost from autonomous hospitals, which again counts as a competing incentive against the reform. Therefore, there might be some degree of veiled resistance against the reform at the Iranian medical universities and even the ministry level. Although resistance against liberalization of public hospitals is not unprecedented even among less-centralized health systems such as Germany’s due to the nature of health that should not be treated such as a commodity (26), the resistance in Iran was due to a different reason; losing power among some authorities.

At the implementation stage, the reform encountered various barriers that made the establishment of real autonomous hospitals almost impossible in Iran. The MOHME struggled to announce the autonomous hospitals’ service charges. This was probably because the ministry knew that it could not afford autonomous hospitals’ bills (27), and/or the insurance organizations had warned the MOHME they could not afford the costs, although they had agreed the reform. Hence, the charges were announced late and then neither the MOHME nor the insurance organizations paid hospitals the amounts they had guaranteed to pay. The MOHME could pay just partially and the insurance organizations paid the autonomous hospitals just as same as regular public hospitals. We had no finding about why the insurance organizations did not pay their committed payments to autonomous hospitals while they had agreed this in advance, but a possible explanation could be that the basic insurance organizations had experienced many changes after signing the agreement and their heads did not and could not pay what their predecessors had committed. Moreover, it is possible that the negotiations between the MOHME and insurance organizations had conducted only at their supreme levels, not tailored first among their experts and middle level managers. The delayed or reduced funding when governments decentralize their facilities and delegate some authority has been reported in similar reforms such as the Indonesian health sector reform (28).

All our other findings listed under process were connected and confirming one another. Poor follow up from the MOHME part was obvious and so the medical universities did not take the implementation serious. We did not examine directly why this was not taken serious by asking direct questions, because we could not access the MOHME’s top managers. However, our other findings might answer this; the MOHME could not afford higher costs of the autonomous hospitals at least at short term. Therefore hospital boards would not have enough motives to hold their meetings regularly due to lack of receiving committed budgets form the MOHME and insurance organizations. Moreover the board’s head, the university chancellor, was reluctant to lead the reform because the reform would limit his/her power, both in terms of managerial authority and financial, due to missing a 5% university share of hospitals’ revenue. Moreover, lack of any external overseer and certain ambiguities in rules and regulations would make the chancellor very cautious in implementation of the reform because he/she would have to take all the responsibilities for the consequences.

Contextual factors also proved as serious obstacles against the reform. Lack of any regulation against physicians’ mobility between public and private hospitals and physicians’ natural and traditional interest in working at both sectors hindered the reform. The public sector’s safety and private income would shape a model of dual practice that is obvious in the Iranian health sector although it could be seen across many other nations (29). Therefore, hospitals could not change all their clinicians into full-time ones as the reform required. Moreover, at some rare specialties changing physicians to full-time would not be appropriate as this can limit patients’ access.

Lack of managerial stability at most public organizations in Iran, seen at hospitals as well, was the other contextual barrier, which slowed down the reform. Most managers were not sure if they could establish the reform at their hospitals during their mission at hospitals and so did not proceed on reforms. Even where the autonomous hospitals reform was starting to proceed, it restarted after new managers came. Lack of managerial stability through insurance organizations, discussed earlier, also would probably have obstructed working of the reform, by delayed and decreased payments to the hospitals.

Strengths/Limitations

This paper had some strengths and weaknesses. Our first phase of study included all Iranian medical universities and all autonomous hospitals, to catch their views on obstacles and barriers of the reform. By this comprehensive sample, we potentially have all barriers and experiences of the reform across all possible locations with their specific context. Nevertheless, we could not interview the MOHME authorities who knew policy issues at the highest level of the health sector. Their views might answer some of our concerns.

Comparison with other reforms

The findings had some similarities and differences with those of at different countries where the World Bank’s autonomy hospitals reform was established. In Pakistan, the reform was proved very political and not straightforward (30). It was run generally based on the donors’ will rather than the government’s own, while in Iran no external will obstructed the reform. In Pakistan, also the reform was altered or obstructed by “street bureaucrats” as was in Iran where insurance organizations, university chancellors and physicians did this. In South-East Asia, the reform has proceeded in Philippines, Indonesia and Vietnam but no significant transformation at their public stewardship has occurred (31), and the reform could not improve efficiency and quality of care (32). In Iran, no study has examined the impact of the reform on hospitals’ clinical indicators. The story was a bit different where the autonomous hospitals were implemented at developed countries. UK’s Foundation Trusts are good example of successful liberalization of public hospitals in order to increase business-like practices (25).

Policy and research implications

Our findings have implications for policy formulation and policy implementation. Liberalization of public hospitals, especially in the settings with a dominant public sector, would be a political reform rather than just a managerial delegation. We saw that the Iranian reform was not supported very well by the MOHME as the main policy-maker and its introducer. The MOHME should have taken into account all its capacities and thought about all necessary tools, such as lobbying the key stakeholders or establishing necessary regulations (say for dual practice), and then decide to implement the reform, or at the other case, cease it totally. The Iranian health insurance organizations as the key stakeholders and the most influential bodies in financing the autonomous hospitals did not stand on their commitment, probably because the MOHME’s initial agreement with them was poor and could not satisfy their benefits. More detail policy formulation issues, such as revising the board structure, should have been approved by all direct and indirect stakeholders. In Iran, the board should not be dominated by the affiliating university, but by the local authorities and the hospital if the reform is really going to be implemented. Instead, a strong overseer at higher level would control any violation of law.

We recommend some areas for further research as well. The reform was agreed by the MOHME and insurance organizations, however when it was the insurance organizations’ turn to pay the hospitals based on the agreed amount, they denied their commitment. The reasons behind such behavior are not clear for us and so could be investigated through further research.

Conclusion

The Iranian autonomous hospitals reform, as a major public sector liberalization attempts, was challenged by different factors such as poor policy formulation where the potential stakeholders’ interest were not met. Even the policymakers at the top levels of the MOHME and the key stake-holders seem to have reached no consensus on the implementation and have moved back from their commitments. Therefore, the medical universities and hospitals were struggled at execution of the reform. All our findings bring us to this conclusion that the stages of policy formulation and policy implementation were performed separately in Iran and were not thought simultaneously and in common with those who formulate a policy and those who have to implement and execute it. In addition, the contextual factors, such as prevalent dual practice among physicians, were not considered by the policy makers in order to alter the policy accordingly.

Ethical considerations

Ethical issues (Including plagiarism, informed consent, misconduct, data fabrication and/or falsification, double publication and/or submission, redundancy, etc.) have been completely observed by the authors.