Ординатура / Офтальмология / Английские материалы / Mechanisms of the Glaucomas_Shields, Tombran-Tink, Barnstable_2008

Fig. 4. Pooled prevalence estimates of blindness and low vision in the USA for different racial groups (from reference (24)).

and whites did not differ in the rates of eye examinations undergone in a given year. However, blacks were 78% more likely to undergo surgery and had 76% higher rates of surgical procedures (20). The investigators concluded that the higher rates of glaucoma surgery among blacks may be due to greater disease severity or a delayed onset of care (20). These factors may also help explain differences in blindness estimates between the two groups.

IMPACT OF NEW TECHNOLOGY

The ongoing development of new technologies, capable of detecting glaucoma earlier in the disease course, may potentially affect disease-prevalence estimates. For example, imaging devices such as the confocal scanning laser ophthalmoscope, the scanning laser polarimeter, or optical coherence tomography are enabling clinicians to detect structural damage at an earlier stage in the disease. Moreover, short-wave-automated perimetry and frequency-doubling perimetry can detect functional deficits up to 5 years before they are detectable on standard automated perimetry. By diagnosing glaucoma earlier in the disease course with these technologies, estimates of glaucoma prevalence are expected to increase substantially.

POLICY IMPLICATIONS

In the USA alone, an estimated $2.5 billion is spent annually caring for patients with glaucoma (29). Of this, $1.9 billion is spent in direct costs and $0.6 billion in indirect costs. A recent study by Lee and colleagues (29) found that the average direct cost of glaucoma treatment ranged from $623 per patient annually for glaucoma suspects or patients with early glaucoma to $2511 per patient annually for patients with end-stage disease. Using pooled data and statistical modeling, Quigley and Vitale (24) showed that whites live approximately 12.8 years with glaucoma and blacks approximately 16.3 years. As US adults live longer and the prevalence of OAG increases, this will likely result in the need to allocate more resources to care for patients with glaucoma. By diagnosing and treating OAG early in the disease course, it may be possible to prevent significant visual impairment in many patients and to reduce the direct and indirect costs associated with caring for patients with advanced disease. Alternatively, this strategy of identifying and treating individuals with OAG early in the disease course requires subjecting a large number of individuals to years of therapy, which could be very costly. Hopefully, with the aid of prevalence and predicted OAG prevalence estimates, clinicians and policymakers can work together to target populations at greatest risk of disease, with the goal of preventing vision loss and minimizing unnecessary treatment.

CONCLUSION

This chapter describes some of the factors that affect OAG prevalence estimates and presents estimates of OAG prevalence generated from population-based crosssectional studies, a meta-analysis of many smaller observational studies, and data from large, nationally representative longitudinal Medicare claims databases. In addition to providing useful information about the overall prevalence of OAG, these studies help elucidate the effect of age and race on OAG prevalence. Accurate estimates of OAG prevalence are important for policymakers to determine the necessary resources to care for individuals with OAG and to help identify, on the basis of demographic profile, which patients are at greatest risk for the disease, so they can be identified early in the disease course, before significant vision loss occurs.

REFERENCES

1.Friedman DS, Wolfs RC, O’Colmain BJ et al. Eye Diseases Prevalence Research Group. Prevalence of open-angle glaucoma among adults in the United States. Archives of Ophthalmology. 122(4):532–8, 2004.

2.He M, Foster PJ, Johnson GJ, Khaw PT. Angle-closure glaucoma in East Asian and European people. Different diseases? Eye. 20(1):3–12, 2006.

3.Foster PJ. The epidemiology of primary angle closure and associated glaucomatous optic neuropathy. Seminars in Ophthalmology. 17(2):50–8, 2002.

4.Tielsch JM, Katz J, Singh K et al. A population-based evaluation of glaucoma screening: the Baltimore Eye Survey. American Journal of Epidemiology. 134:1102–10, 1991.

5.American Academy of Ophthalmology. Preferred Practice Pattern: Primary Open-Angle Glaucoma. San Francisco, CA: American Academy of Ophthalmology; 1996.

6.Lee PP, Feldman ZW, Ostermann J, Brown DS, Sloan FA. Longitudinal prevalence of major eye diseases. Archives of Ophthalmology. 121(9):1303–10, 2003.

7.Tielsch JM, Sommer A, Katz J, Royall RM, Quigley HA, Javitt J. Racial variations in the prevalence of primary open-angle glaucoma. The Baltimore Eye Survey. Journal of the American Medical Association. 266(3):369–74, 1991.

8.Klein BE, Klein R, Sponsel WE et al. Prevalence of glaucoma. The Beaver Dam Eye Study. Ophthalmology. 99(10):1499–504, 1992.

9.Kini MM, Leibowitz HM, Colton T, Nickerson RJ, Ganley J, Dawber TR. Prevalence of senile cataract, diabetic retinopathy, senile macular degeneration, and open-angle glaucoma in the Framingham eye study. American Journal of Ophthalmology. 85(1):28–34, 1978.

10.Quigley HA, West SK, Rodriguez J, Munoz B, Klein R, Snyder R. The prevalence of glaucoma in a population-based study of Hispanic subjects: Proyecto VER. Archives of Ophthalmology. 119(12):1819–26, 2001.

11.Varma R, Ying-Lai M, Francis BA et al. Los Angeles Latino Eye Study Group. Prevalence of open-angle glaucoma and ocular hypertension in Latinos: the Los Angeles Latino Eye Study. Ophthalmology. 111(8):1439–48, 2004.

12.Dielemans I, Vingerling JR, Wolfs RC, Hofman A, Grobbee DE, de Jong PT. The prevalence of primary open-angle glaucoma in a population-based study in the Netherlands. The Rotterdam Study. Ophthalmology. 101(11):1851–5, 1994.

13.Wensor MD, McCarty CA, Stanislavsky YL, Livingston PM, Taylor HR. The prevalence of glaucoma in the Melbourne Visual Impairment Project. Ophthalmology. 105(4):733–9, 1998.

14.Mitchell P, Smith W, Attebo K, Healey PR. Prevalence of open-angle glaucoma in Australia. The Blue Mountains Eye Study. Ophthalmology. 103(10):1661–9, 1996.

15.Leske MC, Connell AM, Schachat AP, Hyman L. The Barbados Eye Study. Prevalence of open angle glaucoma. Archives of Ophthalmology. 112(6):821–9, 1994.

16.Buhrmann RR, Quigley HA, Barron Y, West SK, Oliva MS, Mmbaga BB. Prevalence of glaucoma in a rural East African population. Investigative Ophthalmology & Visual Science. 41(1):40–8, 2000.

17.Shiose Y, Kitazawa Y, Tsukahara S et al. Epidemiology of glaucoma in Japan—a nationwide glaucoma survey. Japanese Journal of Ophthalmology . 35(2):133–55, 1991.

18.Wolffs RC, Borger PH, Ramrattan RS et al. Changing views on open-angle glaucoma: definitions and prevalences—The Rotterdam Study. Investigative Ophthalmology & Visual Science. 41(11):3309–21, 2000.

19.Rudnicka AR, Mt-Isa S, Owen CG, Cook DG, Ashby D. Variations in primary open-angle glaucoma prevalence by age, gender, and race: a Bayesian meta-analysis. Investigative Ophthalmology & Visual Science. 47(10):4254–61, 2006.

20.Ostermann J, Sloan FA, Herndon L, Lee PP. Racial differences in glaucoma care: the longitudinal pattern of care. Archives of Ophthalmology. 123(12):1693–8, 2005.

21.Murdoch IE, Cousens SN, Babalola OE, Yang YF, Abiose A, Jones BR. Glaucoma prevalence may not be uniformly high in all ‘black’ populations. African Journal of Medicine & Medical Sciences. 30(4):337–9, 2001.

22.Rosenberg NA, Pritchard JK, Weber JL et al. Genetic structure of human populations. Science. 298:2381–5, 2002.

23.Congdon N, O’Colmain B, Klaver CC et al. Eye Diseases Prevalence Research Group. Causes and prevalence of visual impairment among adults in the United States. Archives of Ophthalmology. 122(4):477–85, 2004.

24.Quigley HA, Vitale S. Models of open-angle glaucoma prevalence and incidence in the United States. Investigative Ophthalmology & Visual Science. 38(1):83–91, 1997.

25.Quigley HA. Number of people with glaucoma worldwide. The British Journal of Ophthalmology. 80:389–93, 1996.

26.Resnikoff S, Pascolini D, Etya’ale D et al. Global data on visual impairment in the year 2002. Bull World Health Organ. 82(11):844–51, 2004.

27.Oliver JE, Hattenhauer MG, Herman D, et al. Blindness and glaucoma: a comparison of patients progressing to blindness from glaucoma with patients maintaining vision. American Journal of Ophthalmology. 133(6):764–72, 2002.

28.Tielsch JM, Javitt JC, Coleman A, Katz J, Sommer A. The prevalence of blindness and visual impairment among nursing home residents in Baltimore. The New England Journal of Medicine . 332(18):1205–9, 1995.

29.Lee PP, Walt JG, Doyle JJ et al. A multicenter, retrospective pilot study of resource use and costs associated with severity of disease in glaucoma. Archives of Ophthalmology. 124(1):12–9, 2006.

DEFINITIONS OF GLAUCOMA

Through the years, the definition of POAG has changed many times since this disease was described in 1861 by Donders’ collaborator introducing the term glaucoma simplex (1). Until that time, glaucoma was considered to be a disease primarily marked by inflammation. Changes in definitions or criteria for POAG will continue as our knowledge grows. Essentially, the name POAG masks our ignorance of its pathophysiological mechanisms. When we have unraveled all genetic and other causes of POAG, the word “primary” could disappear. POAG these days may be defined as “a disease of retinal ganglion cells, characterized by a structural change in the optic disc that is best described as excavation, and by a typical, slowly progressive loss of function that begins in the mid-peripheral field and expands both toward the center and peripherally”

(2). The excavation is often called glaucomatous optic neuropathy (GON), and we will further use glaucomatous visual field loss (GVFL) to describe its corresponding function loss. Albeit this given definition of POAG as a combination of GON and GVFL seems comprehensive, in daily practice it can be hard to reach its diagnosis. It is difficult to compare scientific reports due to variations and interpretations of glaucoma definitions. Intraocular pressure (IOP) as a risk factor for POAG is used too often to define glaucoma or to screen for it. Half of the POAG cases in population-based studies have a normal IOP ≤21 mmHg (3). Central corneal thickness is also known to influence IOP measurement but is rarely taken into account in glaucoma publications. An analysis of glaucoma therapy studies in peer-reviewed journals between 1996 and 1999 revealed that only 28% included optic disc morphology in their definition of glaucoma, against 100% IOP and 41% visual field examination (4). One purpose of scientific papers is to present them to peers for evaluation and a still common statement that a diagnosis of POAG was made by a glaucoma specialist, without specifying the criteria employed, should be abandoned. Also we feel that the commonly used adjudication by one or two experts in case of conflicting data is a source of bias, when no a priori criteria are mentioned. The lack of consensus on criteria (5) partially explains the variation in prevalence and incidence of glaucoma that we will see later.

One of the positive side-effects of epidemiology is that it sharpens the clinician’s mind for exact and transparent POAG criteria and that it may make doctors more aware of possibly different attitudes in a clinical or research setting (see Table 1). There are many publications on IOP and glaucoma, and it is rarely mentioned that there is about 2 mmHg interobserver variation (standard deviation of the difference) in measuring the IOP with the Goldmann applanator (6). Few studies mention how they exactly performed the IOP measurement with specific equipment. For applanation tonometry, the amount of fluorescein in the tear film, the width of the mires, squeezing the eyelids, and the patient holding his breath will influence the result. There is also a discrepancy in coming to an IOP value in an eye, for example, is one measurement enough, the average of two or three measurements, the median of three, or the first identical value while repeatedly measuring (6)? We concluded that the median of three is best (6). There is also variation in handling the IOPs of the two eyes when classifying a patient.

Only recently criteria were formulated (after 50 years of Goldmann perimetry!) to define progression of GVFL (7–10). It also became clear that determining GON with semi-automated imaging is more reliable than ophthalmoscopy alone (11). A problem in

adhering to a combination of GON and GVFL before a diagnosis of definite glaucoma can be made is that perimetry is a subjective method and may be, especially in older people, an unreliable examination method. Thus there is a need to classify persons with a marked GON who are unable to follow a proper perimetry protocol. We tried in the Rotterdam Study to solve these problems by classifying POAG in definite, probable, and possible glaucoma by a clear algorithm (5). First, we classified GON into no, possible, or probable GON, the latter two according to the 97.5th and 99.5th percentiles of the vertical cup–disc ratio, asymmetry between eyes or minimal rim width in our study population. These were measured semi-automatically or in case of poor pictures estimated through biomicroscopy with a 90D lens. Correction for the area of the optic discs did not influence our GON criteria (5). We defined definite POAG as the presence of GVFL with at least possible GON in the presence of an open angle and no history or signs of secondary glaucoma. In the Rotterdam Study, GVFL stands for a visual field defect found after two suprathreshold screening tests, confirmed by Goldmann perimetry, compatible with glaucoma, and not explained by other neurophthalmologic pathology (12). Probable POAG was based on two possibilities: either the presence of GVFL in the absence of any GON or the absence of GVFL with a probable GON. Possible POAG is the presence of possible GON (one or more of the three earlier mentioned criteria greater than or equal to the 97.5 percentile but less than the 99.5 percentile of a given population) in the absence of either a GVFL or a visual field test.

To make POAG prevalence or incidence studies more comparable, we suggest synchronizing examination techniques, diagnostic criteria, and algorithms to provide both summary data as well as data in 5-year strata, corrected for age and sex, and to publish both cumulative incidences and incidence rates (see below).

STUDY DESIGN (13,14)

Glaucoma studies can be divided in experimental and non-experimental or observational studies. In experimental studies, a group of similar persons (e.g., glaucoma

patients) is divided into subgroups that are assigned to an intervention or that are not. Examples are a randomized clinical trial or etiologic animal research. In nonexperimental or observational studies (“natural experiments”), no interventions are assigned by the investigator. Non-experimental studies can be divided into crosssectional studies or surveys, and prospective or retrospective longitudinal cohort studies. Biases in observational studies can lead to considerable variation in findings from similar studies, and because we focus more on population-based observational studies, we will expand a bit about their advantages and problems.

In case–control studies, patients with a disease (cases) are compared with unaffected persons (controls). Parameters of interest are compared between both groups. This design is valuable when the occurrence of outcome is rare. However, many biases can influence the results. Selection of an appropriate control group can be difficult; the groups should ideally only differ in disease status. Because many case–control studies use retrospective information, recall-bias can occur, mostly causing under-reporting of exposure in control group. Retrospective data can introduce general inaccuracy of information leading often to an under-estimation of associations. Also missing data (e.g., from hospital files) are a problem.

A cross-sectional or prevalence study aims to describe a specific population at a specific time. Prevalence of a disease refers to the number of cases in a given sample or population at a certain time point. Exposure information is ascertained simultaneously with disease information, and all persons are examined only once or several times within a relatively short time span. From prevalence studies, associations between a disease and possible risk factors can be assessed. However, because it is often unknown whether the risk factor was present before or after the disease was diagnosed, not much can be said about causal mechanisms. For a relatively uncommon disease like POAG with a prevalence in a white population around 2%, compared to systemic hypertension with a prevalence of over 30% in the older population, a large sample size is necessary to obtain reliable point estimates.

In a cohort study, a specific group of persons is identified as the cohort and followed in time. In the cohort, several subgroups might be defined based on exposures. Retrospective cohort studies have the advantage of being less time consuming and thus rather cheap. However, obtaining reliable and complete retrospective data is often difficult, partly due to changing examiners, examination methods and equipments, and recall-bias. These problems can be minimized by prospective cohort studies, which are obviously more time consuming. Cohort studies are well suited for studying risk factors and incidences, for example, the number of new cases over a given time span. Incidence can be estimated from a prevalence measurement, using the change of prevalence as a function of age. A proper incidence measurement, however, requires two measurements in the same population at two different points of time. As a result, incidence studies are logistically challenging, especially in a disease like glaucoma, with its slow course and relatively low prevalence, requiring large cohorts to be followed for many years. As a consequence, loss to follow-up (unable to participate at follow-up or death) is a major problem with glaucoma incidence studies. Incidence can be determined either by following a cohort of patients without disease at baseline, yielding a cumulative incidence, or by using a dynamic population, yielding an incidence rate. Cumulative

incidence is the number of new cases divided by the number of persons in the cohort at baseline. A dynamic population is defined as one in which people drop out due to loss to follow-up, non-response or death, or enter the study at a later time point. In this type of population, the incidence rate should be calculated, defined as the number of new cases divided by the number of years people who participated in the investigation, the so-called person years. Population-based POAG incidence studies are by nature of the older populations highly dynamic and thus warrant incidence rates. From these, a cumulative incidences can be calculated (15).

Cohort studies can be population-based or clinic-based. One can avoid low prevalence and statistical power problems by performing a prospective study in a clinic-based setting. This will lead to a large number of POAG cases in a relatively short time span. It may be difficult, however, to obtain sufficient and well-chosen controls. Clinicbased studies often have selection bias, and extrapolation of their data to the general public may be more difficult, especially in countries where a large proportion of the population has limited access to medical care. Because in population studies about 50% of the POAG cases were unknown prior to the study (5,16), this could introduce serious selection bias. An advantage of clinic-based studies is that they may have a better established diagnosis of POAG, because it is easier to have repeated examinations in a clinic than in a population-based study.

A difficulty with long-term follow-up, especially in clinic-based studies, may be changes in examiners, diagnostic routines, equipment, definitions, protocols, and treatments. Clinic-based studies are also often biased (selection bias) toward higher risk participants, resulting in higher incidence rates than population-based studies. Another potential problem in follow-up studies is surveillance bias, when high-risk subjects are examined more often or more carefully. Compare, for example, diabetic persons who have their eyes checked more often for retinopathy than non-diabetic ones. Surveillance bias can be eliminated by examining all subjects each time in an identical way.

Despite the above-mentioned drawbacks, incidence studies are to be preferred over prevalence studies, especially for the purpose of risk-factor analyses. In prevalence studies, risk-factor analyses are limited to whether a simultaneous occurrence of a presumed risk factor and the disease occurs more often than would be expected by chance only. In incidence studies, risk-factor analyses start with the documentation of the presence or absence of presumed risk factors at baseline, that is, in people without the disease. Hence, results from incidence studies are more likely to uncover causality although they cannot prove causality.

PREVALENCE OF PRIMARY OPEN-ANGLE GLAUCOMA

Glaucoma is the second leading cause of blindness worldwide after cataract (17). The number of subjects with open-angle glaucoma (OAG) worldwide in 2010 is estimated at almost 45 million (18). This number is calculated based on several prevalence studies. Numerous studies have been performed, resulting in many different prevalence rates (see Table 2). Diagnostic criteria in these studies vary from precisely described algorithms based on optic nerve head, retinal nerve fiber layer, and visual field characteristics to consensus meetings or even single expert judgment without description of criteria on the other side. The effect of different diagnostic criteria was illustrated