perspectiva atual Cost-effectiveness in glaucoma . Concepts , results and current perspective

Glaucoma is the leading cause of irreversible blindness in Brazil. Its prevalence and incidence tend to increase significantly in the future, mainly due to the population increase and aging. The scarcity of health care resources and the increasing costs in health require a balanced analysis of health interventions and an efficient allocation of resources. The cost-effectiveness and cost-utility studies are important because they allow a comparison between different alternatives in terms of both their costs and their results. For this purpose, mathematical modeling (such as Markov modeling) is commonly used as the analytical method. The literature, including in Brazil, has plenty of evidence of cost-effectiveness and cost-utility in glaucoma. This article aims to review in a practical way the concepts of economic evaluation in health, describe the different types of health economic studies, as well as the results of cost-effectiveness and cost-utility studies in glaucoma in the literature.


U
nderstanding the cost of a certain condition is crucial to the economic planning of efforts aimed at reducing the burden of this condition. 16] Thus, its economic impact is significantly increased.Knowledge of the cost of using resources and standards of treatment of glaucoma patients are important condition to assess the impact of this increase in prevalence in resources for health.
As glaucoma is a disease of genetic origin, its primary prevention is still impracticable and the only way to prevent blindness is through early diagnosis and effective treatment.Often the diagnosis is made with advanced disease, where the necessary resources for treatment and control are generally more costly.In General, a cost increase is observed with increased severity of the disease; that is, the more advanced the disease, the more is spent with it. 7he ageing of the world population, along with increased costs in health and resource scarcity, requires a cost-effective allocation of resources in the treatment and control of glaucoma. 8here is evidence that the costs decrease and the quality of life improves with early diagnosis and treatment in glaucoma. 8A cost-effective analyses has already been used in glaucoma for different goals.
The aim of this study is to have an updated and practical review of the concepts and of the more recent studies involving a cost-effective analysis, a specific type of economic study on health for glaucoma.

Studies of economic evaluation in health
The availability of new forms of treatment and diagnosis requires knowing how to better allocate resources.According to Vianna and Caetano (2001), the health technology assessments (HTA) have become important for a number of reasons: great variability of clinical practice, uncertainty about the true impact of certain diagnostic or therapeutic interventions, speed of development and dissemination of new technologies, and incompatibility between new technologies and the ones already established. 9he health economic evaluations are part of the HTA, where both the costs and the consequences of health technologies, like drugs, devices and medical equipment; medical procedures, behaviors, and models of health care, are assessed in a parameterized way.There are different types of studies proposing an economic evaluation in health.][11][12][13] The main types of economic evaluation differ as to the outcome measure used for the study of the consequences of the alternatives under study.][11][12][13] Whenever the alternatives have the same outcome (final result) and in the same magnitude, it makes no sense to compare the outcome.In such cases, a cost-minimization study is more indicated.In this type of study, the costs of each alternative are collected and compared.][11][12][13] Both the costs and the consequences of the alternatives are assessed in the cost-effectiveness studies, because the outcomes of these alternatives are the same, but in different magnitudes.In this type of study, the effectiveness is measured in natural units (for example: ratio of patients reaching the target intraocular pressure (IOP), or ability to reduce IOP in mmHg).The results are usually expressed by the incremental costeffectiveness ratio (ICER), which shows how much is a unit of added benefit with the alternative under study.][11][12][13] Four results are possible in the cost-effectiveness assessments: the alternative studied is more expensive and more effective; it is more expensive and less effective; it is cheaper and more effective, and it is cheaper and less effective.When the alternative to be assessed is more expensive and less effective, we say that it was "dominated" by the other alternatives under study, and in theory it should be rejected.][12] The doubt is usually present when the alternative is more expensive and more effective, or cheaper and less effective.Should these alternatives be adopted or rejected?To answer that, we use the value of ICER corresponding to the cost of each benefit achieved with each alternate (for example: R$/mmHg of IOP reduction).There is a theoretical limit under which administrators or the society would be willing to pay to reach a particular health benefit.This limit is called the threshold for cost-effectiveness.In the USA, this threshold is usually considered as $50,000/benefit achieved. 12The World Health Organization (WHO) suggests a value of up to 3 times the value of the GDP (Gross Domestic Product) per capita in the country.In Brazil this amount would be approximately R$81,000.00/benefit. 14herefore, an intervention is considered cost-effective if it is the "dominant" alternative (more effective and cheaper at the same time), or if it is more expensive and more effective, but with costs per benefit not exceeding the threshold of cost-effectiveness.
Cost-effectiveness studies should only be used in cases where the outcome measure has a great clinical significance and is a terminal event of the alternatives under study.Otherwise, there is the risk of difficulty in interpreting and generalizing the results.Due to this difficulty in finding good clinical outcome parameters for many pathologies, a metric capable of capturing an important outcome and with clinical significance for any pathology was considered.
All diseases have an impact on the quality of life or the amount of life, or on both.A measure simultaneously encompassing the concept of maximizing the quality and quantity of life of patients was developed from the concept of patient preference (utility values).2][13][15][16] When the outcome measure in studies of economic evaluation is QALY or DALY, the study is named cost-utility study.Some consider this type of study as a subtype of the costeffectiveness study.Anyway, it is the most indicated one, because it allows the results to be compared among different pathologies, because the outcome studied (quality and quantity of life) is the same.The reasoning for the studies of cost-utility is the same as for the cost-effectiveness studies.2][13][15][16] Finally, in cost-benefit studies, both costs and benefits are measured and expressed in monetary values.There are ways to monetize the benefits in health, but they are open to criticism and enhancements.The results of this type of assessment can be expressed by the incremental cost-effectiveness ratio (the same reasoning of ICER), or by calculating the net benefit, which is the difference between the monetary values earned with each alternative and costs (for example: 10 Other less common economic evaluation studies are as follows: cost-consequence studies and disease cost studies.The cost-effect studies try to make an inventory of all the costs and all the results related to an intervention in health, but data is presented in a disaggregated way, and there is no calculation of the ICER.The cost of illness studies try to identify all costs involved in a specific disease (prevention, diagnosis, treatment and rehabilitation costs).These studies may be related to the prevalence of the disease (costs in a given period of the disease) or the incidence of the disease (costs throughout the life of the new cases of the disease). 10n important part in any health economic study is the definition of the types of costs to be studied.There are different types of costs, which are: direct costs (medical and non-medical), indirect costs and intangible costs.Direct medical costs include fees, hospital costs, medicines, equipment, procedures, etc., and direct non-medical costs are those related to transportation, caregiver, rehabilitation, etc.On the other hand, the indirect costs are linked to the loss of productivity of the patient and/or caregiver.Intangible costs (pain, suffering, etc.) are difficult to measure, and are usually excluded from the analysis.][11][12] The health economic evaluation studies increasingly use methods based on decision analysis in their methodology.These methods will be presented below.

Decision analysis on economic evaluation in health
The decision analysis is the application of an analytical method (model) to compare systematically different decision options.It graphically demonstrates the options, and facilitates the calculation of values required to compare them.In the area of health economics, it helps in the selection of the best alternative or more cost-effective one. 11t is important to remember that the quality of the results of a decision analysis is closely linked to the information used to build the model.Decision makers shall assess critically the structure of the model, as well as the probabilities, cost estimates and the assumptions used. 11he most used types of economic modeling in health studies are: decision trees, Markov modeling and simulation of discrete events. 17he decision trees are simpler models involving short-term consequences on health interventions, such as prevention or treatment of acute infections, for example.They can also be used for some chronic diseases that can be cured (for example, for a surgical treatment).] In cases involving chronic or complex diseases presenting changes in the health condition over a period of time, the models of state transition or the Markov model should be the chosen. 11- 12These models enable researchers to identify changes in quality of life, the quantity of life, and the cost of the disease or intervention in a time period when different options of health interventions are studied.2] The probabilities of transition between different health states can be kept constant over time (Markov chain analysis), or may differ depending on the cycle (Markov process analysis). 11The limitation of Markov modeling is to assume that the probability of moving from one state to another does not depend on medical conditions that the patient might have experienced before, which may not be a realistic representation for some researches.] The discrete event simulation models are used when the health condition changes discreetly over time, and therefore are more flexible than Markov modeling.These models represent complex processes and situations due to the events that may occur and may represent different clinical situations, such as: exacerbation of a process, pathological decompensation or a new clinical situation. 17

Outcome measures for economic evaluation in glaucoma
The outcome measures can be different in the study of glaucoma.Measures such as capacity of pressure reduction are the most commonly used in clinical studies, but they have little value for economic evaluation studies.
For cost-effectiveness studies, some outcome measures that could be considered in the economic evaluation would be: reduction of intraocular pressure (mmHg or percentage in relation to the baseline measure); proportion of stable patients during treatment; ratio of patients who reached the target IOP (e.g.< 18 mmHg); ratio of blindness at the end of the follow-up period; etc.
However, as noted, the amount of possible outcomes to be analyzed in a single disease like glaucoma is huge.Hence the need to use an outcome measure, such as QALY (derived metric of utility values), which takes into account the quality of life affected by glaucoma, but still allowing the comparison with different harms to health. 1516] By convention, a utility value equal to 1.0 is related to a perfect health condition, while a utility value of 0.0 is related to death. 9-13, 15-16, 18Utility values enable the objective quantification of the improvement in the quality of life afforded by an intervention, and lead to the composition of QALY.][12] There are mainly 4 ways to obtain the utility values: time trade off (TTO), standard gamble (SG, willingness to pay (WTP) and multi-attributes (utility values obtained indirectly from generic research instruments like the EuroQol-5 D). 19 The TTO, the most used method, first asks the person to answer how much remaining life (in years) they imagine they In evaluating a patient with suspected glaucoma, the study of risk factors is crucial to decide who should be treated.According to Doshi and Singh, from the economic point of view, it would be more cost-effective to treat only those patients at high and moderate risk of developing glaucoma. 8Medeiros and colleagues developed a risk calculator based on the work of Ocular Hypertensive Treatment Study (OHTS) able to estimate the risk of developing glaucoma in 5 years according to the age, IOP, central corneal thickness, aspect of the optic nerve, Pattern Standard Deviation (PSD) index of the computerized campimetry, and presence of Diabetes mellitus. 30Kymes et al. used the Markov model to discover that the treatment of ocular hypertension with a high risk of developing glaucoma (IOP > 24 mmHg and risk of progression > 2%) was cost-effective from the individual point of view, when measured in QALYs. 31They considered that the treatment of all ocular hypertension was not cost-efetivo. 31he same result was found by Stewart et al. 32 These authors suggest selecting and treating only patients at high risk of conversion, for example: old age, high IOP, and low corneal thickness, and increased excavation of the optic nerve.32 Primary Narrow-Angle Glaucoma (PNG) Only one study in the literature has evaluated the costeffectiveness relation in PNG. Inthis Chinese study, the authors showed that the oblique illumination with flashlight and the measure of depth in the peripheral anterior chamber of these eyes provided lower cost per QALY gained among the strategies studied for prevention and detection of this type of glaucoma.There is no evidence in the literature that the cost effectiveness relation of using peripheral iridectomy with laser has been evaluated as a way to prevent the onset and development of primary narrow-angle glaucoma.33 Primary Open-Angle Glaucoma (POAG) The diagnosis and treatment of POAG are normally held in routine appointments in the ophthalmologist office.American researchers have evaluated if this pattern of conduct would be cost-effective.They found that the diagnosis of POAG in the routine consultation followed by the clinical treatment generated an ICER between US$46,000 and US$28,000 per QALY, therefore being cost-effective to American standards. 34nother object of economic evaluation studies in the treatment of POAG is the treatment strategy.What type of treatment is more cost-effective (clinical, laser or surgery)?What is the most cost-effective medication?These are some examples of questions made by researchers in Brazil and in the world.
Noeker and Walt evaluated the cost-effectiveness of glaucoma treatment with monotherapy using one of the three prostaglandin analogues (bimatoprost, latanoprost and travoprost).In this study, they found that the bimatoprost would present the best cost-effectiveness, concluding that in addition to the traditional evaluations of safety and efficacy of medicines, the cost-effectiveness analysis should be considered to assist in individual and collective decision-making. 35] Yu et al compared trabeculectomy as the initial treatment versus the traditional treatment approach, starting with medications and developing for surgery as needed.In a 4-year period, the initial therapy with trabeculectomy was more costeffective for patients with moderate to advanced glaucoma. 39n Australia, the economic impact of glaucoma was evaluated with an economic model.In this study, the authors will still have.Then they must answer how many of these remaining years they believe they have they would be willing to trade (lose) for perfect health until they die.This ratio (remaining years / imaginary life expectancy) is then subtracted from 1.0 to give the utility value. 19n SG, the respondent is asked which is the highest percentage of immediate death risk they would be willing to take for a treatment to bring them perfect health for the rest of their lives.This percentage value is subtracted from 1.0 to give the corresponding utility value. 19he utilities obtained by the WTP method are calculated from the percentage of a certain amount of monetary value (salary, savings, etc.) that the person would be willing to pay to be cured from a disease.This percentage value is subtracted from 1.0 to give the corresponding utility value. 19he Brazilian Ministry of health, by its Guidelines for Economic Evaluation, suggests the use of the TTO method to obtain the utility values for the Brazilian population. 16tility measures have many advantageous features to evaluate the quality of life.They are: they cover several dimensions of the individual as a whole; sensitive to small changes in health condition; reliable and reproducible; applicability to all health areas; short time to be applied; easily understood by the respondents; good internal validity; can be used for studies of health economics. 20] Patients with bilateral amaurosis (absence of light perception in both eyes) have an average utility value of 0.26, whereas patients with perception of light in one eye have an average value of 0.47. 216][27] In Brazil, Paletta Guedes et al. obtained utility values with two methods (SG and TTO) for patients with primary openangle glaucoma.They noted that the utility values were inversely proportional to the visual function in the better eye, whether measured with visual acuity or the Mean Deviation (MD) index of automated perimetry (visual field).The utility values by the TTO to the Brazilian population ranged from 0.84 for visual acuity better than 20/40 to 0.71 for visual acuity worse than 20/ 200.When considering the visual function depending on the degree of commitment of the visual field in the better eye, the utility values (TTO) ranged from 0.86 (MD better than -6.00 dB) to 0.75 (MD worse than -12.00 dB). 22he same authors also evaluated if the type of treatment (surgical or clinical) of glaucoma would influence the utility values for this population.The results of this study showed that the type of treatment, whether clinical, surgical or laser, has no impact on the utility values. 28

Cost-effectiveness and Cost-Utility in Glaucoma
Screening and Treatment of Ocular Hypertension Burr et al. studied if the population screening for glaucoma would be cost-effective.Their findings showed that the screening of the general population was not cost-effective due to the high cost and low prevalence in the general population.However, screening groups considered of high risk for glaucoma could be cost-effective. 29st-effectiveness in glaucoma.Concepts, results and current perspective Rev Bras Oftalmol.2016; 75 (4): 336-41 found that the sequence of laser trabeculoplasty followed by a treatment with clinical eyedrops and surgery, when required, would be more cost-effective than any other treatment sequence.] Stein et al. used the Markov modeling to demonstrate that in the U.S. both the initial treatment with laser and the initial treatment with prostaglandins were cost-effective.These authors found that in a 25-year period the ICUR was US$16,824/QALY for laser and US$14,179/QALY for the clinical treatment. 42n Brazil, the deep non-penetrating esclerectomia was more cost-effective than the prostaglandin analogues alone or in combination with other medications in a 5-year period of follow-up. 43In another study, the surgery of glaucoma (deep non-penetrating esclerectomia) was compared to clinical treatment (3 medications per patient) in a period of 5 years.The results of this study showed that the clinical treatment alternatives were dominated by the surgery, being SUS (Brazilian Single Health System) the perspective of costs. 44The methodologies in these two studies have important limitations preventing or hindering a generalization and practical applicability of their results for the Brazilian population.The effectiveness of interventions was measured in percentage of reduction of the PIO, and not in QALYs.The authors considered a relatively short horizon (5 years) for a chronic illness, and used decision trees rather than the more suitable Markov model.In addition, funding costs were incomplete (consultations, examinations, etc. were excluded).
There is also a shortage of quality scientific evidence of cost-effectiveness and cost-utility for the treatment of glaucoma within the Single Health System (SUS).Such evidence could facilitate decision-making and help establish practice guidelines and clinical conducts the for Reference Centers for the Treatment of Glaucoma from SUS.

CONCLUSION
The analysis of cost-effectiveness and cost-utility is a study of great value in today's world, where there is a shortage of health resources associated to the increased availability and costs of technologies and interventions in health.When including both the costs and the consequences of the alternatives studied in the analysis, the results provide a parameterized reference to assist the health decisionmaking.
The methodology employed is usually based on mathematical models that are fed by the best scientific evidence available about the subject.
There is already enough evidence of cost-effectiveness and cost-utility in the field of glaucoma.However, in Brazil, evidence is still weak, and the studies are incipient.Further studies with rigorous methodology are required to improve the efficiency of the diagnosis and the treatment of glaucoma.