Structural Validity and Reliability of the Healthcare Professionals’ Economic Reasoning Questionnaire

Introduction: The structure of healthcare professionals’ economic reasoning (HPER) is still unknown. The aim: The structural validity and the reliability of the HPER questionnaire were evaluated. Methods: The psychometric study about evaluation of the HPER factors was conducted. The healthcare professionals (physicians, dentists, pharmacists) that are employed in healthcare state sector of Republic of Serbia were interviewed. The HPER Questionnaire contained 29 preliminary items. Its structure was evaluated by Principal Component Analysis for categorial and ordinal data. Results: The 9 items that defined seven HPER factors were extracted: (1) the negligence of the cost of quality by the administration and fi nancier; (2) the consideration of the market price and healthcare benefit; (3) market orientation considering patient’s requests; (4) the recognition of inadequate resource allocations by non-medical administrative authority; (5) the attitude about redirecting the profit from state to private healthcare sector; (6) the recognition of the unjustified spending by the healthcare professionals and (7) the relationship physician/patient. Conclusion: The HPER-9 Questionnaire has high structural validity and reliability in the HPER measurement.


INTRODUCTION
Technical and technological promotions in the healthcare system of the highly developed countries led to the increase of diff erent cost categories in their healthcare economy [1][2][3][4]. State institutions that are responsible for conducting the healthcare economic reforms, reached the consensus that the healthcare professionals' medical decision is evidence based practice, that is the key factor that must be monitored and controlled in order to restrain further cost increase, but to also create the space for further promotion of the citizen healthcare [5]. It is important to highlight that making the medical decision in accordance with the evidence based practice is only a segment of the healthcare professionals' econom-ic reasoning (HPER). From the psycometric, sociometric and econometric point of view, the healtcare professional's reasoning needs to accept the assumption of this problem's multidimensionality [5][6][7][8].
On the other hand, the whole series of the assumed components of the HPER is known, among which there are some that point out: the experience of personal responsibility in the cost control, enthusiasm in creating the strategy to reduce the costs [6], observing professional role in the cost constrain, observing the obstacles and practical consequences of the concerning about the costs [7], awareness and attitude about providing healthcare [8].
However, the assumed psycometric factors of the HPER are practically not competent to be evaluated in some extreme situations, such as the extreme social-economic poverty that is followed by the lack of human resources in healthcare system [9] or some other disorders that are presented by low structural and some other healthcare quality indicator value [10].
Also, it is noticed that the psychometric factors of the economic reasoning actually remained on the assumed level instead of being developed to the level of the evaluated HPER structure. In other words, the assumed structure of a few measurement instruments of the HPER, was not analyzed by psychometric methods that can proof, expand, accept or reject these beforehand assumed structures [5][6][7][8]. On the other hand, the clarifi cation of the changes in the level of the potential HPER factors is very important when planning, conducting and evaluating the outcomes od healthcare-economic and/or healtcare-political reforms [8,11,12].
Having the above mentioned in mind, we considered important to psychometrically analyze factors or dimensions of the HPER.

THE AIM
Th is study's goal is to evaluate the structural validity and reliability of the healthcare professional's economic reasoning questionnaire with defi ning it's latent structure.

Study design
Th e study was conducted as an observational study that covered a cohort of interviewed healthcare professionals that are employed in healthcare institutions of the Republic of Serbia in time period from 2010 -2013 year.

Population
Th e following was included in the study: 1. healthcare professionals (phisicians, dentists or pharmacists); 2. employees of the institutions of the primary, secondary or terciery healthcare or the employees in pharmacy institutuions and 3. that have willingly accepted to take part in the interviewing according to the HPER Questionnaire. Th e following was excluded from the study: 1. the interviewed healthcare professionals that lack more than 10% of data according to the HPER Questionnaire or 2. the interviewed healthcare professionals that have more than 10% of multiple rated answers according to the HPER Questionnaire.

Th e measurement instrument
As a base for the approximate defi nition of the individual 29 items according to the HPER Questionnaire that was applied in our study, [13] we used the HPER Questionnaire (Skootsky AS at al,1999) version that contained 11 items [8]. Modalities of the off ered answers according to the HPER-29 were: I quite agree, I agree, I don't have opinion, I don't agree and I don't agree at all. Th e off ered ordinal anwer modalities will be expressed numerically according to the upward Likert's scale from 0-4, corresponding the level of the agreement with every item or assertion, so that the minimal level of agreement (I don't agree at all) is rated as a 0, whereas the maximum level of agreement (I quite agree) is rated as a 4. Th e interviewing of the healthcare professionals was conducted in their spare time lasting 10 minutes.

Statistical methods
Th e description of the measurement variables was done by showing the mean as a measure of the central tendention of the set of data, while the variation of the set of data was shown by the standard deviation. Th e structural validity and reliability with defi ning the latent structure of HPER-19 and data reduction, was evaluated by Principal component analysis for categorical and ordinal data (CATAPCA). In the process of the elimination of the items and extraction of the dimensions (latent components) in CATAPCA model, it was accepted www.hophonline.org the total explained variance of minimum 85%, with minimal reliability shown by the Crombach's coeffi cient α of 0.7 for each extracted dimension. Th e accepted level of the statistical signifi cance was 0.05. To process the data we used the "SPSS 15" (Chicago, IL).

RESULTS
Th e study was conducted in 66 healthcare institutions of the Republic of Serbia. Out of 1574 310 Volume 2 • Number 3 • October 2015 • HOPH included healthcare professionals, the 102 interviewees were excluded from the study due to the unacceptable number of missing data, so that the total number of healthcare professionals that was analyzed by the structure of the economic reasoning was 1472.
With CATAPCA model with nine items it resulted in seven-dimensional structure that defi ned 90.786 % of the total variance (Table 1). Each dimension resulted in high Cronbach's α coeffi cient (Table 2), while Table 1. CATAPCA model -Factors loading per items and dimensions of the HPER-9 Questionnaire D1 -The negligence of the cost of quality by the administration and fi nancier D2 -The consideration of the market price and healthcare benefi t D3 -Market orientation considering patient's requests D4 -The recognition of inadequate resource alocations by non-medical administrative authority D5 -The attitude about redirecting the profi t from state to private healthcare sector D6 -The recognition of the unjustifi ed spending by the healthcare professionals D7 -The relationship physician/patient * The item recoding is done to obtain particular score (0 becomes 4, 1 becomes 3, 2 stays 2, 3 becomes 1, and 4 becomes 0) • Dimension 1 score (D1 score) -Th e negligence of the cost of quality by the administration and fi nancier is expressed as a mean value of the recoded values for the items 4 and 5: D1 score = (i4 recoded + i5 recoded)/2. • Dimension 2 score (D2 score) -Th e consideration of the market price and healthcare benefi t is expressed as a mean value for the the total Cronbach's α was 0.988. Th e HERP-9 Questionnaire and seven highly structurally valid and reliable dimension were obtained. Descriptive statistics for dimension scores and total score are presented in the Table 3.
Calculation of the scores for dimensions and total HPER-9 score obtained by the CATAPCA model (Table 1): items 1 and 2: D2 score = (i1 + i2)/2. • Dimension 3 score (D3 score) -Market orientation considering patient's requests is expressed as a mean value for the items 7 and 9: D3 score = (i7 + i9)/2. • Dimension 4 score (D4 score) -Th e recognition of inadequate resource alocations by nonmedical administrative authority is expressed as a mean value for the items 3 and 8: D4 score = (i3 + i8)/2 • Dimension 5 score (D5 score) -Th e attitude about redirecting the profi t from state to private healthcare sector is expressed as a mean value of the recoded value for the item 6 and non-recoded value for the item 9: D5 score = (i6 recoded + i9)/2. • Dimension 6 score (D6 score) -Th e recognition of the unjustifi ed spending by the healthcare professionals is expressed as a mean value of the tem 3 and recoded value of the item 8: D6 score= (i3 + i8 recoded)/2 www.hophonline.org • Dimension 7 score (D7 score) -Th e relationship physician/patient is expressed as the value of the item 7: D7 score = i7 • Th e total score of the economic reasoning of health professionals (Total score of the HPER-9), is calculated the following way: Total score of the HPER-9 = (i1 + i2 + i4 recoded + i5 recoded + i6 r recoded + i7 + i8 + i8 recoded + i3 × 2 + i9 × 2).
Th e minimal value of this score (worst economic reasoning) is 4 (total of i8 and i8 recoded is always 4), and maximal score value (best economic reasoning) is 44. As this scale has an absolute zero that amounts 4, while the maximum value is 11 times bigger than the absolute zero, we obtained new continuing scale of the HERP measurement.

DISCUSSION
Skootsky et al 1999 defi ned the assumed HPER structure as the healthcare professionals' attitude for: 1. the concept of providing healthcare and 2. the cost containment policy [8].
However, these authors redefi ne this beforehead assumed structure fenomenologically and concetually as an important, but not enoguh clarifi ed segment of the physicians' professional satisfaction.
In our earlier studies, the assumed structure of our HPER-29 measurement instrument was defi ned by dimensios that the healthcare professionals use to express: 1. the attitude towards the choice between the alternatives; 2. the attitude towards the quality of the healthcare and 3. the attitude towards the cost control policy [13].
Aft er the extensive data reduction in this study we have shown that there is a latent, seven-dimensional HPER-9 Questionnaire structure.
Th e fi rst factor is defi ned by the level of the agreement with the statements "...administration and fi nancial policy in healthcare are considerably involved in the doctor / patient relationship..." and "... administration and fi nancial policy in healthcare are more focused on cost savings than with the provision of quality health care...". We named this factor Th e negligence of the cost of quality by the administration and fi nancier. From the nomenclature and cost quality classifi cation point of view, through the level of the agreement with these items, the healthcare professionals show the level of the cost recognition that are generated by the administration and the fi nanciers while neglecting the healthcare quality in the institutions. In economics, these costs are known as nonconformance cost of quality [14,15]. Since the higher level of the cost quality negligence means lower economic healthcare eff ectivness, answers to these items 312 Volume 2 • Number 3 • October 2015 • HOPH are recoded in order for scale to get the opposite direction, or, in other words, for the higher score to express less cost quality negligence (better economical heathcare eff ectiveness). It's interesting that this factor, compared to the others in our seven-dimensional structure, has the highest variance and reliability explained ( Table 2). Th e latter is important, considering that the concept of the healthcare cost quality hasn't been defi ned yet, which is one of the reasons why they are not shown as extracted items in the regular fi nancial calculations in healthcare institutions [14]. We also highlight that, so far, a small number of studies have been published that contained the subject of the mutual connection of costs and healthcare quality [16][17][18][19]. Th e recognition and the quality cost control is extremely hard and complex job for the management in healthcare institution to do, especially in the conditions where they are limited to do it by the fi nanciers. On the other hand, the eff ect of the management in the cost quality control is carefully observed by the healthcare professionals in the institution, which makes that the Th e negligence of the cost of quality the most important HPER factor.
Since the second factor is determined by the items "...I am familiar with market price of all alternative interventions (preventive, diagnostic test, curative or rehabilitating)..." and "...I am familiar with cost-eff ectiveness ratio for each alternative considered...", this factor is defi ned as Th e consideration of the market price and healthcare benefi t. It corresponds with making the evidence based practice medical decision, due to which this is a known and very important HPER facor [5].
Th e third factor is defi ned by the level of the agreement with the items "...the physicians employed within State owned Institutions don`t have the same level of the dedication to the patients as the physicians in the private sector do..." and "... it seems to me that I have observed an irrational recommendation of procedures lacking fi rm evidence on effi ciency..." Th is factor expresses the signifi cance of the devotion to the patient whose requests need to be considered and acknowledged whenever possible. Since, at the same time, the patient is both the recipient of the healthcare services and the participant in the fi nancing through the health insurance, this factor is named Market orientation considering patient's requests.
Th e fourth factor is defi ned by the level of the agreement with the items "...it seems to me that I have observed an irrational recommendation of procedures lacking fi rm evidence on effi ciency..." and "...I would avoid prescribing medicines strictly based on instructions of the non-medical management body..." Th is factor expresses the enthusiasm for the healthcare professional to express themselves about the unjustifi ed spending, but also about the bad economic infl uences of the nonmedical administrative authority due to the lack of or the bad bidirectional communication between the healthcare professionals and the nonmedical administrative authority. Th at's why we named this factor Th e recognition of inadequate resource alocations by non-medical administrative authority. Th e connection establishment of the healthcare administration and management with the healthcare professionals, that encurages the enthusiasm of the healthcare professionals to express themselves about the unjustifi ed spendings in the healthcare institution, lately becomes important in the better developed countries. Th is becomes important because of the more effi cient control and evaluation of the eff ects of the current economic healthcare reforms in these countries [20]. In other words, by observing this factor there can be considered or even taken the appropriate measures in order to prevent the negative infl uence of the pharmaceutic corporations to the behaviour of the nonmedical administrative authority, and also of other corporations whose interest is to keep or to enlarge the placement of the products with small net benefi ts.
Th e fi ft h factor is defi ned by the level of the agreement with the items "...physicians employed within State owned facilities due to strict inner and outer control, run fewer tests and examinations per patient compared to the physicians employed within privately owned facilities..." and "... if the patient demands certain laboratory analysis or the imaging examination and the risk of harm is acceptable, the patient should be given the service demanded..." Since this factor expresses the attitude that the state institutions compared to the private ones cannot make the profi t by mostly answering the patients' needs and requests, we named this factor Th e attitude about redirecting the profi t from state to private healthcare sector. Since the item 6 ( Table  1) expresses bad economic eff ects (from the www.hophonline.org least bad to the most bad), we recoded this answer scale too, so that the scale could have the ascendant character (from the lowest to the highest economic eff ectiveness). Th e extraction of this factor is important from the legal and economic point of view, considering that in the United States of America a physician in profi t-oriented institutions, when concluding the contract of employment, must accept the clause that aft er leaving the institution will not conduct the medical practice in the same or a nearby community [21. Since there are clear evidence that this kind of requests are strictly profi t-motivated, these clauses were seriously criticized with an explanation that they can potentially harm the patients by limiting their rights to freely chose their physician [22]. Th e healthcare professionals' observation of the profi t redirection from state to private sector can be a signifi cant indicator of the strengthening of the enterprenureal behaviour in private sector, or the shortage of the enterprenureal behaviour in the state sector.
Th e sixth factor is defi ned by the level of the agreement with the items "...it seems to me that I have observed an irrational recommendation of procedures lacking fi rm evidence on effi ciency... " and "... I would avoid prescribing medicines strictly based on instructions of the non-medical management body... " Th is factor, too, expresses the enthusiasm of the healthcare professional to freely express themselves about the unjustifi ed spending that, opposite to the fourth factor, do not come from the nonmedical administrative authority (because the corresponding factorial coeffi cient has a negative sign -see Table1). We named this factor Th e recognition of the unjustifi ed spending by the healthcare professionals. Answers to the item 8 are recoded, due to the negative factorial coeffi cient, so that the scale gets the appropriate (opposite) direction.
Last, the seventh factor is defi ned by answering one item "...the physicians employed within State owned Institutions don`t have the same level of the dedication to the patients as the physicians in the private sector do..." Since this item highlights only the signifi cance of the devotion to the patient and the signifi cance of the quality of the established relationship between the physician and the patient, we named this factor Th e relationship physician / patient. Th is factor expresses the attitude of the healthcare professionals that is shaped by legal, cultural and ethical norms, such as patients' le-gal rights, medical ethical code etc. Also, this factor defi nes the appliability of the previously mentioned regulations as a refl ection of the level of the quality culture according to the dimensions of the equity, patient centeredness and the timelines of health service delivery. In our earlier studies, we showed that the patient's satisfaction mostly refl ects the achieved levels in two healthcare quality dimensions: 1. patient centeredness of medical staff and 2. the timeliness of health service delivery [23].

CONCLUSION
Considering the content and the sense of the factors that together explain, sinthesize and that way defi ne the studied phenomenology, we can defi ne the measurement of the Healthcare Professionals' Economic Reasoning as a measurement of the subjectively expressed level of the healthcare professionals' experience of their own economic role, management economic role and the economic role of other healthcare professionals in the institution. Since we extracted the obtained factor with the CATAPCA modeling, this means that the mutual relations between the items in the HPER-9 Questionnaire are not linear and that the specifi c items will have, by participating in one factor (or dimension), completely diff erent meaning when being a part of any other factor.