Magnitude of eye injuries worldwide
Unlike other major blinding disorders such as cataract, trachoma, onchocerciasis or xerophthalmia, where epidemiological studies have contributed significantly to a better understanding of disease patterns, in the case of ocular injuries epidemiological data are scarce or totally lacking for large parts of the world. In fact, eye injuries have been considered a clinical issue, and are mostly addressed within the context of clinical eye care delivery systems including emergency case management. However, like any other eye disorder, eye injuries do not occur as random events: there is evidence that some population groups are at increased risk because of greater exposure to hazards, decreased ability to avoid or detect hazards, and/or a lower likelihood of functional recovery following eye injury.
From a public health point of view, it is the epidemiology of injury, not of accidents that is most important. The scientific interest of all these disorders is their potential for producing permanent anatomical or functional complications (e.g., disfigurement, visual impairment).
The purpose of this article is to estimate the magnitude of eye injuries and their consequences at the global level, utilising available data both from the Most Developed Countries (MDCs) and Less Developed Countries (LDCs). Risk factors and the pattern of eye injuries will not be discussed in this paper. The present article is based on work carried out in the WHO Programme for the Prevention of Blindness and Deafness as part of the development of the Global Data on Blindness.
According to Gibson,1 an injury is damage to a person or to a tissue/organ, e.g., the eye, caused by a transfer of energy, namely one of the five forms of physical energy: mechanical, thermal, chemical, electrical and radiant. Consequently, eye injuries include all damage caused to the eye and its adnexa, orbital and periorbital tissues due to direct contact with fixed or mobile, blunt or sharp objects (mechanical energy being transferred for example by rapid deceleration), hot objects (thermal), chemical substances, sources of electrical power, different types of radiation (UV, X-ray, microwave).
Sources of information
- A primary source of information is found in the existing ophthalmic literature. Garrow2 reported 1000 consecutive cases of eye injury registered in Scotland between 1908-1913, however, from this date, publications of reliable epidemiological data remain relatively scarce in the ophthalmic literature. Much of the information available is essentially descriptive, and mainly based on hospital data, i.e., without clear reference to a defined population. Most published studies are based on data provided by outpatient departments, hospital admission/discharge forms, or emergency departments.
- Another source of information is the WHO Programme for the Prevention of Blindness (PBL) which has carried out or supported a number of population-based, cross-sectional random sample surveys on blindness and its causes (including eye trauma) in several countries (Table 1).
- In order to assess the magnitude and severity of eye injuries from an epidemiological perspective, Négrel and Thylefors3 have recently selected and reviewed 33 hospital-based case series to estimate risk factors and clinical outcomes, 11 population-based studies to estimate parameters such as incidence and prevalence, and 10 cross-sectional random sample surveys to estimate the magnitude of low vision/blindness in relation to eye trauma. Much of this material will be presented in this article.*
* The full article, The Global Impact of Eye Injuries, will be published shortly in Ophthalmic Epidemiology. For further information about this publication, please contact Aeolus Press, PO Box 740, 4116 ZJ Buren, The Netherlands
Selection of possible indicators
In most parts of the world it is impossible, through a literature review, to assess the epidemiological parameters (incidence or prevalence) of eye injuries. Generally, it is possible to approach this issue through indirect indicators such as the share (as proportion/percentage) of eye injuries in the total workload of a specific health care centre. The incidence of eye injuries restricting activities or requiring a ‘contact’ with an eye care centre during a defined period of time, or through the prevalence of binocular or monocular visual loss due to eye injuries, can be recorded.
(1) Percentage of eye injuries in relation to total attendance at healthcare centres
Analysis of this type of data3 confirms that eye injuries represent a very common cause of presentation and admission at health centres, although the figures vary largely from one study to another: 10% to 27% of all cases examined in outpatient departments; 38% to 65% of all the cases seen in emergency departments, and 5% to 16% of all admissions in eye hospitals are related to eye injuries. As a proportion of all hospital admissions for ocular disorders, eye injuries are becoming relatively more important, in part as a result of the recent shift of many ophthalmic procedures, such as outpatient cataract surgery, to outpatient settings.
(2) Incidence of eye injuries restricting activity, or requiring a contact with a health facility
In assessing the frequency of eye injury from a public health point of view, those requiring medical attention are of prime importance because of the direct costs of the care and the time lost from productive activity. A prospective observational study of cases of moderate to severe trauma, carried out in Scotland during one year has recently been reported. The overall incidence of ocular trauma, requiring hospitalisation under the care of a consultant ophthalmologist, was 8.14 per 100 000 of the population in one year.4
Table 1. Prevalences (per 100 000) of blindness/low vision* due to eye trauma. Review of 10 cross-sectional random sample studies (Source WHO/PBL/94.40: Available Data on Blindness – An update – unpublished data) *
|Country||Year||Population examined||Blindness due to eye injury||Low vision due to eye injury||Monocular loss of vision due to eye injury **|
* Blindness is defined as vision <3/60. Low vision is defined as vision <6/18 but ³3/60.
** Includes monocular blindness and monocular low vision.
From the reviewed literature,3 the following estimates can be considered as a reasonable model for industrialised countries, in order to assess the magnitude of the problem worldwide
- Incidence of eye injury restricting activity for more than one day 900 to 1000/100 000 population/year
- Incidence of eye injury requiring medical attention 400 to 1 000/100 000 population/year
- Incidence of eye injury requiring hospitalisation 8 to 13/100 000 population/year
The incidence of ‘acute hospital-treated eye injury’ can vary dramatically in a relatively short time. In recent years, for example, the development of the chemical industry in rapidly industrialising countries has resulted in higher incidences of injuries caused by chemicals. On the other hand, the introduction of effective preventive measures, such as the required use of car seat belts or protective eye-wear in amateur ice hockey, has resulted in an impressive reduction of severe eye injuries necessitating hospitalisation.
Unfortunately, reliable estimates of incidence in developing countries are not available.
(3) Prevalence of eye injuries
Prevalence is routinely used to measure the total number of existing cases found in a defined population at one point in time. Considering that the majority of eye injuries are clinically insignificant or mild, or remain without complications, and are of short duration, relatively very few cases of eye injury are found during cross-sectional surveys which provide population-based information.
In an attempt to evaluate the magnitude of the problem, another estimate of prevalence could be used, namely ‘cumulative prevalence’ (the total number of persons who have suffered from any type of eye injury at any time during a specified period which, unfortunately, depends largely on recall, and is more likely to be biased). Two important studies, not entirely comparable, deserve to be reported here as examples, one for the MDCs, the other for LDCs:
(i) The Baltimore Eye Survey3 in the USA examined the cumulative lifetime prevalence of ocular injuries that included the spectrum of mild, visually non-threatening to severe blinding injuries in a multiracial urban population, aged 40 and over. In this study, at least one injury in a lifetime was reported by 22.5% of black men, 20.3% of white men, 12.2% of black women and 7.7% of white women.
(ii) The Nepal Eye Study6 provided a prevalence estimate for the whole population. Among the 39 887 persons examined in the survey, 336 were found to have signs and history of previous eye injury. Prevalence was estimated at 860/100 000 population. There was a marked increase in prevalence with increasing age. Prevalence rose from 340/100 000 for persons under age 10, to 1780/100 000 for persons aged from 55 to 59. This study also demonstrated that 62% of injured persons had signs of eye injury but no visual impairment, 27% were unilaterally blind or severely visually impaired, 8% were bilaterally visually impaired and 3% were blind.
Table 2. Global estimates of eye trauma and blindness/low vision/monocular blindness in LDC’s and MDC’s
|Less Developed Countries 1996 population: 4 633 447 a||More Developed Countries 1996 population:1 170 673 a||Total 1996 population: 5 804 120 a|
|Estimated rate (per 100 000)||Estimated number||Estimated rate (per 100 000)||Estimated number||Estimated number|
|Yearly incidence of eye injuries restricting activities for more than one day||–||(44 000 000) b||950||11 000 000||(55 000 000) b|
|Yearly incidence of open-eye injury.||–||(162 000) b||3.5||41 000||(203 000) b|
|Yearly incidence of eye injuries necessitating hospitalisation||–||(600 000) b||13||150 000||(750 000) b|
|Blindness||30 c||1 500 000||9||100 000||1 600 000|
|Low vision||45 c||2 100 000||13.5 d||(160 000)||2 260 000|
|Monocular blindness||330 c||15 300 000||300||3 500 000||18 800 000|
a United Nations Demographic Yearbook 1996: figures in thousands
b Global extrapolation on assumption of similar rates between Less- and More Developed Countries
c Rounded-off weighted mean estimated from ten selected studies
d Estimated rate of low vision, based on the assumption of a multiplying factor (×1.5) between blindness and low vision
(4) Prevalence of blindness, bilateral low vision and unilateral visual impairment in relation to eye injuries
Visual impairment and blindness are most often the result of complications from eye injury (such as secondary infections or sympathetic ophthalmia), but blindness may also be due to the immediate severity of the injury itself.
(i) For industrialised countries, survey results, published 20 years ago by the US National Society for the Prevention of Blindness7 stated that
The prevalence of blindness due to injury was 9/100 000
The prevalence of monocular blindness was estimated to be around 300/100 000
(ii) For developing countries, 10 well conducted, population-based cross-sectional studies have been reviewed.3 These studies, presented in Table 1, indicate that:
- Prevalence estimates of blindness due to eye injury range from 0 to 75 per 100 000
- Prevalence estimates of bilateral low vision due to eye injury range from 30 to 137 per 100 000
- Prevalence estimates of unilateral visual impairment due to eye injury range from 0 to 490 per 100 000
These indicators have been weighted and then incorporated in the model proposed below in an attempt to assess the approximate global magnitude and severity of eye morbidity due to eye injury.
1. Assessment of the Magnitude of Eye Morbidity in Relation to Eye Injuries
It is clear that the information available is very scarce, and direct comparisons between studies remain difficult. Therefore, the development of a model to assess the global magnitude of eye injuries, becomes a very uncertain exercise.
(i) For MDCs, it is possible to develop an algorithm which incorporates reasonable values from the following indicators:
- the incidence of eye injuries restricting activities for more than one day
- the incidence of eye injuries requiring hospitalisation
- the incidence of open-eye injuries
(ii) For LDCs, no such information on ‘incidence’ could be found.
(iii) The first step in the model should be to separate these two broad groups of countries (MDCs and LDCs). The model (as used in Table 2) assumes that the epidemiological patterns for eye injuries are similar in countries within a given group.
The demographic data reported in Table 2 are taken from the UN Demographic Yearbook for 1996. Where data are not available, extrapolation is proposed on the assumption that similar rates are observed in LDCs and in MDCs. (The corresponding figures are displayed within brackets). It was thought that applying rates from MDCs to LDCs allows reasonable minimum estimates to be proposed. Using the weighted parameters from the literature review gives the following estimates to be proposed for 1996:
- An estimated global incidence of eye injuries of 950/100 000 population restricting activities for more than one day; this implies an annual incidence of 55 million injuries of this kind
- An estimated global incidence of eye injuries of 13/100 000 population requiring hospitalisation; the overall yearly incidence for such injuries would be around 750 000 cases
- An estimated global incidence for open-eye injuries of 3.5/100 000 population which indicates more than 200 000 such cases per year.
In our opinion, the reported figures represent a gross underestimate of the actual situation for the following reasons:
- Many patients with minor eye injuries never seek eye care
- Patients who sustain eye trauma but seek care elsewhere than the hospital – private practice offices, ambulatory care centres – are not included in the published statistics
- Patients who develop late complications, such as cataract, retinal detachment, secondary glaucoma, and are diagnosed months or years after the event, are often registered differently than under the label of ‘acute eye injury’ or trauma
- For many patients sustaining an eye injury in the context of severe multiple trauma, the ocular lesions are often considered as a ‘secondary diagnosis’, and consequently are under-represented in the published series
- One of the major drawbacks of hospital-based statistics is the assessment of the actual size of the population to which the trauma sample belongs
2. Assessment of blindness and severely impaired vision in relation to eye injuries
The amount of blindness due to eye injuries in the world has never been accurately estimated. The last update of the Global Data on Blindness stated that ‘other causes to consider include ocular trauma, estimated to be responsible for about 500 000 cases of blindness’ In 1992, Thylefors8 reported that, in developing countries, ocular trauma represents about 5% of all blindness cases.
Population-based studies, available mostly for the LDCs, provide the basis for estimating the consequences of ocular injuries in terms of visual loss However, there are wide variations in the prevalence rates (Table 1). To overcome this, a weighted mean was taken and rounded off The estimated rate of low vision is based on the assumption of a multiplying factor of 1.5 between blindness and low vision estimated after analysing the published results of a compilation of ten population-based surveys (Table 1)
For MDCs, the estimates provided by the US National Society for the Prevention of Blindness are included in the proposed model, even though they are more than 20 years old Considering the absence of a better estimator, the rate of low vision for MDCs was reached by applying the same proportionate relationship of 1 5 times the blindness rate calculated for LDCs
Using the definitions of the International Classification of Diseases for low vision and blindness, the following current global estimates were arrived at in this way (Table 2)
- A total of 1 6 million cases of blindness are caused by eye injuries
- Some 2 3 million cases with low vision are due to eye injuries
- Some 19 million cases of monocular blindness are due to eye injuries
These estimated figures suggest that eye injuries represent a major public health problem It is also true that they represent more ‘the cause of a blind eye rather than the cause of a blind person’. Nevertheless, uniocular injuries may be very disabling, and occur most frequently in the active years of life, and consequently the vocational and economic consequences are enormous.
In developing countries, eye injuries are not only more common but also more severe in their effects. Since eye injuries are related to particular occupations or cultural environments, the type and prognosis of injuries seen in developing countries are not similar to those in industrialised countries. In Africa, and in many parts of Asia, eye injuries present their own patterns, not only in terms of aetiology or severity, but also in relation to the socioeconomic background, reflecting the non-existence or inadequacy of safety measures, the lack of proper eye health facilities to provide adequate case management, the use of traditional medicines, poor education, and a lack of awareness amongst manual workers in hazardous occupations. The prognosis of the initial lesions is commonly made worse by delay before proper management, or by the administration of inappropriate home medication. In fact, five major factors are linked to the final outcome for the injured eye: (i) the severity of the initial lesion (especially in the case of an open-eye injury), (ii) the first aid treatment provided, (iii) the time elapsed from injury to definitive care, (iv) the quality of care, (v) the pre-existing eye health status (e.g., myopia, wear of contact lenses).
Considering the scarcity of available data, and inconsistency in the way these are reported, there have been some assumptions made in the use of the algorithm which need further validation.
Based on the model presented, estimates of the incidence of eye injuries and related visual impairment and blindness prevalences significantly exceed those previously published.
There is an urgent need for more consistent recording of eye injury events and their consequences, to provide more accurate data for planning eye care services. In addition, such data would also provide better understanding of the changing patterns of eye injuries in different parts of the world.
The author wishes to express thanks to Dr Pararajasegaram and Dr Thylefors for their assistance in reviewing this manuscript.
1 Gibson JJ. The contribution of experimental psychology to the formulation of safety: a brief basic research. In: Behavioural approaches to accident research. New York, Association for the Aid of Crippled Children, 1961.
2 Garrow A. A statistical inquiry of 1000 cases of eye injuries. Br J Ophthalmol 1923; 7: 65-80.
3 Négrel AD, Thylefors B. The Global Impact of Eye Injuries. Ophthalmic Epidemiology. (54 bibliographic references). In press.
4 Desai P, MacEwen CJ, Baines P, Minassian DC. Incidence of cases of ocular trauma admitted to hospital and incidence of blinding outcome. Br J Ophthalmol 1996; 80: 592-6.
5 Katz J, Tielsch JM. Lifetime prevalence of ocular injuries from the Baltimore eye survey. Arch Ophthalmol 1993; 111: 1564-8.
6 Trauma in the epidemiology of blindness in Nepal. In: Brilliant GE, eds. Report of the 1981 Nepal Blindness Survey. San Rafael, CA. The SEVA Foundation, 1988.
7 National Society to Prevent Blindness (1980). Vision problems in the US. Facts and figures.
8 Thylefors B. Epidemiological patterns of ocular trauma. Aust NZ J Ophthalmol 1992; 20(2): 95-8.