- Email: [email protected]
The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey
G Model JINJ 7421 No. of Pages 4
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Full length article
The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey Stuart Keela,* , Jing Xiea , Joshua Foremana,b , Hugh R. Taylorc , Mohamed Dirania a
Centre for Eye Research Australia, Royal Victorian Eye & Ear Hospital, Melbourne, Australia Ophthalmology, University of Melbourne, Department of Surgery, Melbourne, Australia c Indigenous Eye Health Unit, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 6 September 2017 Accepted 19 September 2017
Background: To determine the prevalence of vision loss due to ocular trauma in Australia. Methods: The National Eye Health Survey (NEHS) is a population-based cross-sectional study that examined 3098 non-Indigenous Australians (aged 50–98 years) and 1738 Indigenous Australians (aged 40–92 years) living in 30 randomly selected sites, stratified by remoteness. An eye was considered to have vision loss due to trauma if the best-corrected visual acuity was worse than 6/12 and the main cause was attributed to ocular trauma. This determination was made by two independent ophthalmologists and any disagreements were adjudicated by a third senior ophthalmologist. Results: The sampling weight adjusted prevalence of vision loss due to ocular trauma in non-Indigenous Australians aged 50 years and older and Indigenous Australians aged 40 years and over was 0.24% (95% CI: 0.10, 0.52) and 0.79% (95% CI: 0.56, 1.13), respectively. Trauma was attributed as an underlying cause of bilateral vision loss in one Indigenous participant, with all other cases being monocular. Males displayed a higher prevalence of vision loss from ocular trauma than females in both the non-Indigenous (0.47% vs. 1.25%, p = 0.03) and Indigenous populations (0.12% vs. 0.38%, p = 0.02). After multivariate adjustments, residing in Very Remote geographical areas was associated with higher odds of vision loss from ocular trauma. Conclusions: We estimate that 2.4 per 1000 non-Indigenous and 7.9 per 1000 Indigenous Australian adults have monocular vision loss due to a previous severe ocular trauma. Our findings indicate that males, Indigenous Australians and those residing in Very Remote communities may benefit from targeted health promotion to improve awareness of trauma prevention strategies. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Epidemiology Public health Ocular injury
Introduction Ocular injuries are a leading, but avoidable, cause of monocular vision loss globally [1]. In Australia, there are an estimated 20,000 hospitalisations due to ocular injury annually, [2] at a direct cost of $155 million [3]. Indigenous Australians [2], males [4] and individuals residing in rural areas [5–7] have consistently been reported to be at a higher risk of ocular trauma. Despite the notable public health concern, recent population-based data on the frequency of severe ocular trauma in Australian adults remains limited. Most data on ocular injury in Australia comes from hospitalbased reports [2,3,8,9]. To date, the most robust population-based
* Corresponding author at: Level seven, Centre for Eye Research Australia, 32 Gisborne Street, East Melbourne, Victoria, Australia. E-mail address: [email protected] (S. Keel).
data can be derived from the Melbourne Visual Impairment Project (VIP) [4] and the Blue Mountains Eye Study (BMES) [10] conducted in the early 19900 s that reported the prevalence of monocular vision loss (<6/12) due to ocular trauma to be 0.25% and 0.33%, respectively. More recent data from the National Indigenous Eye Health Survey (NIEHS, 2008) [11] and the Central Australian Ocular Health Study (CAOHS, 2010) [12] suggest that the burden of ocular injury is notably higher amongst Indigenous Australian adults, with nearly one-third of all monocular blindness attributed to injury. This is in line with national hospital-based data that have reported three-fold higher rates of hospitalised ocular injuries amongst Indigenous Australians than non-Indigenous Australians [2]. Herein, we describe the prevalence and causes of vision loss due to ocular trauma in a national, population-based sample of nonIndigenous and Indigenous Australian adults aged 50 years or older and 40 years or older, respectively.
http://dx.doi.org/10.1016/j.injury.2017.09.020 0020-1383/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
2
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Materials and methods Study population The NEHS is a population-based, cross-sectional survey conducted between the 11th of March 2015 and 18th of April 2016. Details of the sampling methodology have been described in detail elsewhere [13]. Multistage random-cluster sampling was used to select thirty Australian sites across five Remoteness Areas (RAs) that included Major City, Inner Regional, Outer Regional, Remote and Very Remote geographical areas, based on data from the 2011 Australian Census [14]. The Australian Bureau of Statistics assigns a remoteness classification to a Statistical Area according to the Accessibility/Remoteness Index of Australia (ARIA+) classification system. ARIA+ scores range from 0 to 15 along a continuous scale based on road distances between Statistical Areas and their nearest service centres (e.g. ARIA+ of >10.53–15.00 = Very Remote). Indigenous Australians aged 40 years or older and non-Indigenous Australians aged 50 years or older were recruited door-to-door (examination rate = 71.5%). The younger age criteria for Indigenous participants was chosen due to the earlier onset and more rapid progression of common eye diseases and diabetes in Indigenous Australians [15], coupled with a lower life expectancy [16]. The protocol was approved by the Royal Victorian Eye and Ear Hospital Human Research Ethics Committee (HREC-14/1199H) as well as State-based Indigenous organisations. Study procedures adhered to the tenets of the Declaration of Helsinki as revised in 2013 and participants provided written informed consent to participate. Examination procedures The examination protocol of the NEHS has been described in detail elsewhere [17]. Socio-demographic data including age, gender, Indigenous status, ethnicity, years of education and language spoken at home were collected via an intervieweradministered questionnaire. Participants self-reported whether they had ever been told that they have cataracts, age-related macular degeneration (AMD), diabetic retinopathy (DR), glaucoma. Further to this, participants self-reported whether they had any ‘other’ ocular history, including ocular or adnexal (e.g. eyelids) trauma, and if so they were prompted for the details about the cause. Presenting distance visual acuity (VA) was measured in each eye using a logMAR chart (Brien Holden Vision Institute, Australia) in well-lit room conditions. Pinhole testing was performed on participants with visual acuity worse than 6/12 in one or both eyes, followed by automated refraction (Nidek ARK-30 Type-R Handheld auto-refractor/keratometer, Nidek Co., LTD, Japan) if VA improved to 6/12 or better in either eye. Examination of the anterior segment was performed using a hand-held slit lamp (Keeler Ophthalmic Instruments, UK) at 10 magnification. Participants with VA worse than 6/12 in either or both eyes had anterior segment photographs taken of the impaired eye(s) using a Digital Retinography System (DRS) camera (CenterVue SpA, Italy).
Two-field, 45 colour fundus photographs were taken of each retina using the DRS camera, centred on the macula and optic disc, respectively, in a darkened room to allow for physiological mydriasis. De-identified images were transferred to the retinal image grading centre at the Centre for Eye Research Australia (CERA), where blinded retinal graders graded images according to protocols that have been described in detail elsewhere [18–20]. An eye was considered to have visual impairment due to trauma if best-corrected distance VA was <6/12-6/60 due to trauma. If VA was worse than 6/60 due to trauma, this was considered to be blindness due to trauma. The term vision loss incorporated both visual impairment and blindness, defined as a best-corrected VA of <6/12. The determination of whether the loss was due to trauma was made by two independent ophthalmologists who reviewed relevant questionnaire and clinical data. Any disagreements were adjudicated by a third senior ophthalmologist. Statistical analysis Bivariate descriptive statistics were utilized to test for statistically significant differences in all study variables between participants who experienced ocular injuries versus those who did not. Chi-square tests for homogeneity were used to test for differences in categorical variables; two-tailed Student's t-tests were used to test for differences in continuous variables. The occurrence of ocular injury is a rare event. The main issue of logistic regression with the rare event data is a serious bias problem in regression coefficient estimates. So the sampling weight-adjusted prevalence rates of vision loss from ocular injury were estimated using Poisson distribution for rare event, stratified by Indigenous status. Poisson regression was used to assess differences in the prevalence rates of vision loss from ocular injury by gender, age, education, and geographic location, mutually adjusting for each of these factors in a multivariable model. Incidence rate ratios (IRR) and their estimated 95% confidence intervals (CI) were estimated from the Poisson regression analyses. All analyses were performed by incorporating the sampling weights and non-response rates to obtain unbiased estimates from the complex NEHS sampling design. Analyses were conducted in Stata 14.2.0 (StataCorp). A P < 0.05 (two-tailed) was deemed statistically significant. Results A total of 4836 individuals were examined in the NEHS, including 3098 non-Indigenous and 1738 Indigenous Australians, respectively. The sample of non-Indigenous Australians had a mean age of 66.6 years (SD = 9.7 years) while the mean age of Indigenous participants was 55.0 years (SD = 10 years). NonIndigenous Australians were 46% male and Indigenous Australians were 41% male. Self-reported history of ocular trauma Overall, 63 (2.0%) non-Indigenous Australians aged 50 years or over and 70 (4.0%) Indigenous Australians aged 40 years or over
Table 1 Prevalence [(% (95% CI)] of vision loss from ocular injury, stratified by Indigenous status. Non-Indigenous (n = 3098)
Total Female Male
Indigenous (n = 1738)
n
Crude% (95% CI)
Weighted% (95% CI)
n
Crude% (95% CI)
Weighted% (95% CI)
10 2 8
0.32 (0.17, 0.60) 0.12 (0.03, 0.48) 0.56 (0.28, 1.11)
0.24 (0.10, 0.52) 0.47 (0.10, 2.10) 1.25 (0.58, 2.70)
12a 3 9
0.69 (0.39, 1.22) 0.29 (0.09, 9.08) 1.26 (0.66, 2.42)
0.79 (0.56, 1.13) 0.12 (0.02, 0.59) 0.38 (0.16, 0.94)
CI = Confidence Interval. a 1 Indigenous Australian had binocular vision loss from ocular injury.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
characteristics (Table 2). After adjusting for covariates, residing in a Very Remote geographical area (IRR = 3.4, p = 0.02) was associated with higher odds of vision loss from ocular trauma.
recalled having an ocular injury at some time in the past. Of these, 44 (69.8%) non-Indigenous Australians and 52 (74.3%) Indigenous Australians reported the cause of ocular injury. In the nonIndigenous population, the top five self-reported ocular injuries were corneal or conjunctival foreign bodies (21/63, 33.3%), blunt trauma (9/63, 14.3%), gardening (5/63, 7.9%), assault/firearm (4/63, 6.4%) and chemical burn (2/63, 3.2%). For Indigenous Australians, the five leading types of injuries were corneal or conjunctival foreign body (20/70, 28.6%), blunt trauma (18/70, 21.4%), chemical burn (4/70, 5.7%), sports (4/70, 5.7%) and assault/firearm (4/70, 5.7%). Of all participants with a self-reported eye trauma, 33.8% (48/142) had observable structural damage on slit lamp examination.
Discussion Ocular trauma remains an important cause of monocular vision loss in Australia, with our data showing that approximately 2.4 per 1000 non-Indigenous and 7.9 per 1000 Indigenous Australians have vision loss due to previous trauma. Residing in Very Remote geographical areas was associated with higher odds of vision loss from ocular trauma. The prevalence of vision loss due to ocular trauma in nonIndigenous Australian adults appears to have remained relatively stable over the past two decades, with rates in the NEHS similar to that reported in the BMES (0.33%) [10] and the MVIP (0.25%) [4]. In line with previous clinical2 and population-based data [4,11,12], we report 2–3 times higher rates of vision loss from ocular trauma among Indigenous Australians than in non-Indigenous Australians (0.79% vs. 0.24%) and in men compared to women (nonIndigenous; 0.47% vs. 1.25%, Indigenous; 0.12% vs. 0.38%). While these gender differences were not significant after multivariate adjustments (Table 2), this was most likely attributable to the small sample size. The observed higher burden of major ocular trauma in the Indigenous population and males may be explained by a number of well-recognised reasons including; greater occupational and physical risks (e.g. farming or mining work) [21], higher rates of risky health behaviours [22] and lower rates of eye care services utilisation after sustaining trauma [23]. The higher prevalence of vision loss from ocular trauma in individuals residing in Very Remote regions of Australia concurs with national hospital data [5]. While it appears that this finding is largely driven by the Indigenous population (Prevalence; Very Remote = 2.1% vs. Major City = 0.31%), we could not explicitly verify this in regression analysis due to the low frequency ocular trauma causing vision loss in this study. It may be speculated that the observed regional difference could be explained by poorer access to eye health services in Very Remote areas of Australia [24], with a greater lag time between injury occurrence and care seeking placing these individuals at a higher risk of vision loss. Strengths of the current study include the population-based sample that was stratified by Indigenous status. There are also some notable limitations. First, the small number of cases of vision loss from ocular trauma limits the power of this study to detect associations with risk factors. Furthermore, we did not utilise a structured questionnaire to collect specific information related to ocular trauma, including the eye affected, timing, number of episodes, cause, location and whether protective correction was worn at the time of trauma. As a result, we could not report the prevalence and risk factors of any ocular trauma in Australia.
Prevalence of vision loss due to ocular trauma The weighted prevalence of vision loss caused by ocular trauma among non-Indigenous Australians aged 50 years or over was 0.24% (95% CI: 0.10, 0.52) (Table 1). All cases were monocular, including with 2 participants with visual impairment and 8 participants with blindness due to trauma. Among Indigenous Australians aged 40 years and over, the weighted prevalence of vision loss from ocular trauma was 0.79% (95% CI: 0.56, 1.13), consisting of 3 participants with monocular visual impairment, 8 participants with monocular blindness and 1 participant with binocular blindness due to trauma. Males displayed a higher prevalence of vision loss from ocular trauma in the non-Indigenous (1.25% (95% CI: 0.58, 2.70) vs. 0.47% (95% CI: 0.10, 2.10), p = 0.03) and Indigenous populations (0.38% (95% CI: 0.16, 0.94) vs. 0.12% (95% CI: 0.02, 0.59), p = 0.02). Extrapolating these findings to the Australian population, we estimate that approximately 14,395 non-Indigenous Australians aged 50 years or over and 1063 Indigenous Australians aged 40 years or over have vision loss caused by ocular trauma. Among the 10 cases of monocular vision loss from ocular trauma in the non-Indigenous population, 3 eyes (30%) had traumatic cataract, 3 (30%) eyes were enucleated due to previous trauma, 2 (20%) had traumatic corneal scars and 2 (20%) had traumatic retinal detachment or scarring. Trauma resulting in optic atrophy was attributed as an underlying cause of bilateral blindness in one Indigenous participant. Of the remaining 11 cases of monocular vision loss from ocular trauma in the Indigenous population, 3 eyes (27.3%) had traumatic cataract, 3 (27.3%) eyes were enucleated due to previous trauma, 2 (18.2%) had traumatic corneal scars and 2 (18.2%) had traumatic retinal detachment or scarring. Due to the low frequency of ocular trauma causing vision loss found in the NEHS, Indigenous and non-Indigenous data were combined in Poisson regression analysis examining associations between ocular trauma causing vision loss and selected
Table 2 Relationship between ocular injury causing vision loss and selected characteristics in Australians (n = 4836). Associated factors
Indigenous Age (per 10 years) Gender (male) Education Remoteness Major city Inner Regional Outer Regional Remote Very Remote
Univariate logistic regression
Multivariable logistic regression
IRR [95% (CI)]
p
IRR [95% (CI)]
p*
3.26 (1.37, 7.75) 0.18 (0.84, 1.20) 1.14 (0.58, 2.88) 0.05 (0.02, 0.13)
0.01 0.73 0.19 0.18
3.26 (0.61, 17.41) 1.33 (0.43, 4.08) 2.96 (0.48, 18.33) 1.07 (0.97, 1.18)
0.16 0.61 0.23 0.17
1 0.63 (0.07, 5.75) 2.08 (0.39, 10.97) 2.16 (0.51, 9.03) 4.60 (1.54, 13.78)
0.68 0.38 0.28 0.01
1 0.61 (0.07, 5.09) 2.09 (0.38, 11.58) 2.24 (0.55, 9.08) 3.44 (1.24, 9.54)
0.63 0.39 0.25 0.02
IRR = incidence rate ratio; CI = Confidence Interval.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
4
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Rather our data is limited to reporting the prevalence of vision loss due to ocular trauma (based on relevant clinical data). It is also possible that survivor bias could have led to an underestimation in the true prevalence of vision loss due to ocular trauma in this study, where people who experienced severe ocular or head injuries during their lifetime may have been at a higher risk of earlier mortality. In summary, the NEHS provides Indigenous-specific prevalence estimates of vision loss due to ocular trauma in a national population-based sample of Australian adults. The majority of vision loss from trauma is preventable and it impacts a significant number of Australians in their productive years. Our data may indicate that males, Indigenous Australians and those residing in Very Remote communities may benefit from improved health promotion efforts focussing on injury prevention, including the use of protective eyewear. Further research on the annual incidence, causes and location of occurrence of any ocular trauma in Australia would be helpful to design these targeted programs. Conflicts of interest Nil to declare. Acknowledgements The Centre for Eye Research Australia (CERA) and Vision 2020 Australia wish to recognise the contributions of all the NEHS project steering committee members (Professor Hugh Taylor, Dr Peter van Wijngaarden, Jennifer Gersbeck, Dr Jason Agostino, Anna Morse, Sharon Bentley, Robyn Weinberg, Christine Black, Genevieve Quilty, Louis Young and Rhonda Stilling) and the core CERA research team who assisted with the survey field work (Joshua Foreman, Pei Ying Lee, Rosamond Gilden, Larissa Andersen, Benny Phanthakesone, Celestina Pham, Alison Schokman, Megan Jackson, Hiba Wehbe, John Komser and Cayley Bush). Furthermore, we would like to acknowledge the overwhelming support from all collaborating Indigenous organisations who assisted with the implementation of the survey, and the Indigenous health workers and volunteers in each survey site who contributed to the field work. The National Eye Health Survey was funded by the Department of Health of the Australian Government, and also received financial contributions from Novartis Australia and the Peggy and Leslie Cranbourne Foundation. In-kind support was received from our industry and sector partners, OPSM, Carl Zeiss, Designs for Vision, the Royal Flying Doctor Service, Optometry Australia and the Brien Holden Vision Institute. We would like to specifically acknowledge OPSM, who kindly donated sunglasses valued at $130 for each study participant. The Centre for Eye Research Australia receives Operational Infrastructure Support from the Victorian Government. The Principal Investigator, Dr Mohamed Dirani, is supported by a NHMRC Career Development Fellowship (#1090466). The PhD
student, Joshua Foreman is supported by an Australian Postgraduate Award scholarship. References [1] Negrel AD, Thylefors B. The global impact of eye injuries. Ophthalmic Epidemiol 1998;5(3):143–69. [2] Australian Institue of Health and Welfare. Eye related injuries in Australia. 2009. . Accessed 3rd of March 2017 http://www.aihw.gov.au/WorkArea/ DownloadAsset.aspx?id=6442458828. [3] Fong LP. Eye injuries in victoria: Australia. Med J Aust 1995;162(2):64–8. [4] McCarty CA, Fu CL, Taylor HR. Epidemiology of ocular trauma in Australia. Ophthalmology 1999;106(9):1847–52. [5] Australian Institue of Health and Welfare. Australian hospital statistics 2004– 05. Canberra, Australia. Accessed 27th March 2017. 2016. http://www.aihw. gov.au/WorkArea/DownloadAsset.aspx?id=6442456820:2006. [6] Northey LC, Bhardwaj G, Curran S, McGirr J. Eye trauma epidemiology in regional Australia. Ophthalmic Epidemiol 2014;21(4):237–46. [7] Thompson CG, Griffits RK, Nardi W, Tester MP, Noble MJ, Cottee L, et al. Penetrating eye injuries in rural new south wales. Aust N Z J Ophthalmol 1997;25(1):37–41. [8] Casson RJ, Walker JC, Newland HS. Four-year review of open eye injuries at the Royal Adelaide Hospital. Clin Exp Ophthalmol 2002;30(1):15–8. [9] Smith AR, O'Hagan SB, Gole GA. Epidemiology of open- and closed-globe trauma presenting to Cairns Base Hospital, Queensland. Clin Exp Ophthalmol 2006;34(3):252–9. [10] Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia: the Blue Mountains Eye Study. Ophthalmology 1996;103(3):357–64. [11] Taylor HR, Xie J, Fox S, Dunn RA, Arnold AL, Keeffe JE. The prevalence and causes of vision loss in indigenous australians: the national indigenous eye health survey. Med J Aust 2010;192(6):312–8. [12] Landers J, Henderson T, Craig J. The prevalence and causes of visual impairment in indigenous Australians within central Australia: the Central Australian Ocular Health Study. Br J Ophthalmol 2010;94(9):1140–4. [13] Foreman J, Keel S, Dunn R, van Wijngaarden P, Taylor HR, Dirani M. Sampling methodology and site selection in the National Eye Health Survey (NEHS): an Australian population-based prevalence study. Clin Exp Ophthalmol 2016. [14] Australian Bureau of Statistics. Australian statistical geography standard (ASGS). 2014. . Accessed 2nd April 2017 http://www.abs.gov.au/websitedbs/ D3310114. nsf/home/Australian+Statistical+Geography+Standard+(ASGS). [15] Australian Institute of Health and Welfare. Vision problems among older Australians. 2005. . Accessed 10th February 2017 http://www.aihw.gov.au/ WorkArea/DownloadAsset.aspx?id=6442453390. [16] Australian Institute of Health and Welfare. Mortality and life expectancy of indigenous Australians. 2014. . Accesed 10th of February 2017 http://www. aihw.gov.au/WorkArea/DownloadAsset.aspx?id=60129548468. [17] Foreman J, Keel S, van Wijngaarden P, Taylor HR, Recruitment Dirani M. Testing protocol in the national eye health survey: a population-based eye study in Australia. Ophthalmic Epidemiol 2017;1–11. [18] Happich M, Reitberger U, Breitscheidel L, Ulbig M, Watkins J. The economic burden of diabetic retinopathy in Germany in 2002. Graefes Arch Clin Exp Ophthalmol 2008;246(1):151–9. [19] Ferris F, Wilkinson C, Bird A, Chakravarthy U, Chew E, Csaky K, et al. Clinical classification of age-related macular degeneration. Ophthalmol 2013;120:844–51. [20] Mukesh B, McCarty C, Rait J, Taylor H. Five-year incidence of open-angle glaucoma. Ophthalmol 2002;109:1047–51. [21] Berecki-Gisolf J, Smith PM, Collie A, McClure RJ. Gender differences in occupational injury incidence. Am J Ind Med 2015;58(3):299–307. [22] Australian Institute of Health and Welfare. Health behaviours and other risks to health. 2014. . Accessed 21st April 2017 http://wwwaihwgovau/australiashealth/2014/health-behaviours. [23] Arnold AL, Busija L, Keeffe JE, Taylor HR. Use of eye care services by indigenous Australian adults. Med J Aust 2011;194(10):537–8. [24] Kiely PM, Chakman J. Optometric practice in australian standard geographical classification-remoteness areas in Australia, 2010. Clin Exp Optom 2011;94 (5):468–77.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Full length article
The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey Stuart Keela,* , Jing Xiea , Joshua Foremana,b , Hugh R. Taylorc , Mohamed Dirania a
Centre for Eye Research Australia, Royal Victorian Eye & Ear Hospital, Melbourne, Australia Ophthalmology, University of Melbourne, Department of Surgery, Melbourne, Australia c Indigenous Eye Health Unit, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 6 September 2017 Accepted 19 September 2017
Background: To determine the prevalence of vision loss due to ocular trauma in Australia. Methods: The National Eye Health Survey (NEHS) is a population-based cross-sectional study that examined 3098 non-Indigenous Australians (aged 50–98 years) and 1738 Indigenous Australians (aged 40–92 years) living in 30 randomly selected sites, stratified by remoteness. An eye was considered to have vision loss due to trauma if the best-corrected visual acuity was worse than 6/12 and the main cause was attributed to ocular trauma. This determination was made by two independent ophthalmologists and any disagreements were adjudicated by a third senior ophthalmologist. Results: The sampling weight adjusted prevalence of vision loss due to ocular trauma in non-Indigenous Australians aged 50 years and older and Indigenous Australians aged 40 years and over was 0.24% (95% CI: 0.10, 0.52) and 0.79% (95% CI: 0.56, 1.13), respectively. Trauma was attributed as an underlying cause of bilateral vision loss in one Indigenous participant, with all other cases being monocular. Males displayed a higher prevalence of vision loss from ocular trauma than females in both the non-Indigenous (0.47% vs. 1.25%, p = 0.03) and Indigenous populations (0.12% vs. 0.38%, p = 0.02). After multivariate adjustments, residing in Very Remote geographical areas was associated with higher odds of vision loss from ocular trauma. Conclusions: We estimate that 2.4 per 1000 non-Indigenous and 7.9 per 1000 Indigenous Australian adults have monocular vision loss due to a previous severe ocular trauma. Our findings indicate that males, Indigenous Australians and those residing in Very Remote communities may benefit from targeted health promotion to improve awareness of trauma prevention strategies. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Epidemiology Public health Ocular injury
Introduction Ocular injuries are a leading, but avoidable, cause of monocular vision loss globally [1]. In Australia, there are an estimated 20,000 hospitalisations due to ocular injury annually, [2] at a direct cost of $155 million [3]. Indigenous Australians [2], males [4] and individuals residing in rural areas [5–7] have consistently been reported to be at a higher risk of ocular trauma. Despite the notable public health concern, recent population-based data on the frequency of severe ocular trauma in Australian adults remains limited. Most data on ocular injury in Australia comes from hospitalbased reports [2,3,8,9]. To date, the most robust population-based
* Corresponding author at: Level seven, Centre for Eye Research Australia, 32 Gisborne Street, East Melbourne, Victoria, Australia. E-mail address: [email protected] (S. Keel).
data can be derived from the Melbourne Visual Impairment Project (VIP) [4] and the Blue Mountains Eye Study (BMES) [10] conducted in the early 19900 s that reported the prevalence of monocular vision loss (<6/12) due to ocular trauma to be 0.25% and 0.33%, respectively. More recent data from the National Indigenous Eye Health Survey (NIEHS, 2008) [11] and the Central Australian Ocular Health Study (CAOHS, 2010) [12] suggest that the burden of ocular injury is notably higher amongst Indigenous Australian adults, with nearly one-third of all monocular blindness attributed to injury. This is in line with national hospital-based data that have reported three-fold higher rates of hospitalised ocular injuries amongst Indigenous Australians than non-Indigenous Australians [2]. Herein, we describe the prevalence and causes of vision loss due to ocular trauma in a national, population-based sample of nonIndigenous and Indigenous Australian adults aged 50 years or older and 40 years or older, respectively.
http://dx.doi.org/10.1016/j.injury.2017.09.020 0020-1383/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
2
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Materials and methods Study population The NEHS is a population-based, cross-sectional survey conducted between the 11th of March 2015 and 18th of April 2016. Details of the sampling methodology have been described in detail elsewhere [13]. Multistage random-cluster sampling was used to select thirty Australian sites across five Remoteness Areas (RAs) that included Major City, Inner Regional, Outer Regional, Remote and Very Remote geographical areas, based on data from the 2011 Australian Census [14]. The Australian Bureau of Statistics assigns a remoteness classification to a Statistical Area according to the Accessibility/Remoteness Index of Australia (ARIA+) classification system. ARIA+ scores range from 0 to 15 along a continuous scale based on road distances between Statistical Areas and their nearest service centres (e.g. ARIA+ of >10.53–15.00 = Very Remote). Indigenous Australians aged 40 years or older and non-Indigenous Australians aged 50 years or older were recruited door-to-door (examination rate = 71.5%). The younger age criteria for Indigenous participants was chosen due to the earlier onset and more rapid progression of common eye diseases and diabetes in Indigenous Australians [15], coupled with a lower life expectancy [16]. The protocol was approved by the Royal Victorian Eye and Ear Hospital Human Research Ethics Committee (HREC-14/1199H) as well as State-based Indigenous organisations. Study procedures adhered to the tenets of the Declaration of Helsinki as revised in 2013 and participants provided written informed consent to participate. Examination procedures The examination protocol of the NEHS has been described in detail elsewhere [17]. Socio-demographic data including age, gender, Indigenous status, ethnicity, years of education and language spoken at home were collected via an intervieweradministered questionnaire. Participants self-reported whether they had ever been told that they have cataracts, age-related macular degeneration (AMD), diabetic retinopathy (DR), glaucoma. Further to this, participants self-reported whether they had any ‘other’ ocular history, including ocular or adnexal (e.g. eyelids) trauma, and if so they were prompted for the details about the cause. Presenting distance visual acuity (VA) was measured in each eye using a logMAR chart (Brien Holden Vision Institute, Australia) in well-lit room conditions. Pinhole testing was performed on participants with visual acuity worse than 6/12 in one or both eyes, followed by automated refraction (Nidek ARK-30 Type-R Handheld auto-refractor/keratometer, Nidek Co., LTD, Japan) if VA improved to 6/12 or better in either eye. Examination of the anterior segment was performed using a hand-held slit lamp (Keeler Ophthalmic Instruments, UK) at 10 magnification. Participants with VA worse than 6/12 in either or both eyes had anterior segment photographs taken of the impaired eye(s) using a Digital Retinography System (DRS) camera (CenterVue SpA, Italy).
Two-field, 45 colour fundus photographs were taken of each retina using the DRS camera, centred on the macula and optic disc, respectively, in a darkened room to allow for physiological mydriasis. De-identified images were transferred to the retinal image grading centre at the Centre for Eye Research Australia (CERA), where blinded retinal graders graded images according to protocols that have been described in detail elsewhere [18–20]. An eye was considered to have visual impairment due to trauma if best-corrected distance VA was <6/12-6/60 due to trauma. If VA was worse than 6/60 due to trauma, this was considered to be blindness due to trauma. The term vision loss incorporated both visual impairment and blindness, defined as a best-corrected VA of <6/12. The determination of whether the loss was due to trauma was made by two independent ophthalmologists who reviewed relevant questionnaire and clinical data. Any disagreements were adjudicated by a third senior ophthalmologist. Statistical analysis Bivariate descriptive statistics were utilized to test for statistically significant differences in all study variables between participants who experienced ocular injuries versus those who did not. Chi-square tests for homogeneity were used to test for differences in categorical variables; two-tailed Student's t-tests were used to test for differences in continuous variables. The occurrence of ocular injury is a rare event. The main issue of logistic regression with the rare event data is a serious bias problem in regression coefficient estimates. So the sampling weight-adjusted prevalence rates of vision loss from ocular injury were estimated using Poisson distribution for rare event, stratified by Indigenous status. Poisson regression was used to assess differences in the prevalence rates of vision loss from ocular injury by gender, age, education, and geographic location, mutually adjusting for each of these factors in a multivariable model. Incidence rate ratios (IRR) and their estimated 95% confidence intervals (CI) were estimated from the Poisson regression analyses. All analyses were performed by incorporating the sampling weights and non-response rates to obtain unbiased estimates from the complex NEHS sampling design. Analyses were conducted in Stata 14.2.0 (StataCorp). A P < 0.05 (two-tailed) was deemed statistically significant. Results A total of 4836 individuals were examined in the NEHS, including 3098 non-Indigenous and 1738 Indigenous Australians, respectively. The sample of non-Indigenous Australians had a mean age of 66.6 years (SD = 9.7 years) while the mean age of Indigenous participants was 55.0 years (SD = 10 years). NonIndigenous Australians were 46% male and Indigenous Australians were 41% male. Self-reported history of ocular trauma Overall, 63 (2.0%) non-Indigenous Australians aged 50 years or over and 70 (4.0%) Indigenous Australians aged 40 years or over
Table 1 Prevalence [(% (95% CI)] of vision loss from ocular injury, stratified by Indigenous status. Non-Indigenous (n = 3098)
Total Female Male
Indigenous (n = 1738)
n
Crude% (95% CI)
Weighted% (95% CI)
n
Crude% (95% CI)
Weighted% (95% CI)
10 2 8
0.32 (0.17, 0.60) 0.12 (0.03, 0.48) 0.56 (0.28, 1.11)
0.24 (0.10, 0.52) 0.47 (0.10, 2.10) 1.25 (0.58, 2.70)
12a 3 9
0.69 (0.39, 1.22) 0.29 (0.09, 9.08) 1.26 (0.66, 2.42)
0.79 (0.56, 1.13) 0.12 (0.02, 0.59) 0.38 (0.16, 0.94)
CI = Confidence Interval. a 1 Indigenous Australian had binocular vision loss from ocular injury.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
characteristics (Table 2). After adjusting for covariates, residing in a Very Remote geographical area (IRR = 3.4, p = 0.02) was associated with higher odds of vision loss from ocular trauma.
recalled having an ocular injury at some time in the past. Of these, 44 (69.8%) non-Indigenous Australians and 52 (74.3%) Indigenous Australians reported the cause of ocular injury. In the nonIndigenous population, the top five self-reported ocular injuries were corneal or conjunctival foreign bodies (21/63, 33.3%), blunt trauma (9/63, 14.3%), gardening (5/63, 7.9%), assault/firearm (4/63, 6.4%) and chemical burn (2/63, 3.2%). For Indigenous Australians, the five leading types of injuries were corneal or conjunctival foreign body (20/70, 28.6%), blunt trauma (18/70, 21.4%), chemical burn (4/70, 5.7%), sports (4/70, 5.7%) and assault/firearm (4/70, 5.7%). Of all participants with a self-reported eye trauma, 33.8% (48/142) had observable structural damage on slit lamp examination.
Discussion Ocular trauma remains an important cause of monocular vision loss in Australia, with our data showing that approximately 2.4 per 1000 non-Indigenous and 7.9 per 1000 Indigenous Australians have vision loss due to previous trauma. Residing in Very Remote geographical areas was associated with higher odds of vision loss from ocular trauma. The prevalence of vision loss due to ocular trauma in nonIndigenous Australian adults appears to have remained relatively stable over the past two decades, with rates in the NEHS similar to that reported in the BMES (0.33%) [10] and the MVIP (0.25%) [4]. In line with previous clinical2 and population-based data [4,11,12], we report 2–3 times higher rates of vision loss from ocular trauma among Indigenous Australians than in non-Indigenous Australians (0.79% vs. 0.24%) and in men compared to women (nonIndigenous; 0.47% vs. 1.25%, Indigenous; 0.12% vs. 0.38%). While these gender differences were not significant after multivariate adjustments (Table 2), this was most likely attributable to the small sample size. The observed higher burden of major ocular trauma in the Indigenous population and males may be explained by a number of well-recognised reasons including; greater occupational and physical risks (e.g. farming or mining work) [21], higher rates of risky health behaviours [22] and lower rates of eye care services utilisation after sustaining trauma [23]. The higher prevalence of vision loss from ocular trauma in individuals residing in Very Remote regions of Australia concurs with national hospital data [5]. While it appears that this finding is largely driven by the Indigenous population (Prevalence; Very Remote = 2.1% vs. Major City = 0.31%), we could not explicitly verify this in regression analysis due to the low frequency ocular trauma causing vision loss in this study. It may be speculated that the observed regional difference could be explained by poorer access to eye health services in Very Remote areas of Australia [24], with a greater lag time between injury occurrence and care seeking placing these individuals at a higher risk of vision loss. Strengths of the current study include the population-based sample that was stratified by Indigenous status. There are also some notable limitations. First, the small number of cases of vision loss from ocular trauma limits the power of this study to detect associations with risk factors. Furthermore, we did not utilise a structured questionnaire to collect specific information related to ocular trauma, including the eye affected, timing, number of episodes, cause, location and whether protective correction was worn at the time of trauma. As a result, we could not report the prevalence and risk factors of any ocular trauma in Australia.
Prevalence of vision loss due to ocular trauma The weighted prevalence of vision loss caused by ocular trauma among non-Indigenous Australians aged 50 years or over was 0.24% (95% CI: 0.10, 0.52) (Table 1). All cases were monocular, including with 2 participants with visual impairment and 8 participants with blindness due to trauma. Among Indigenous Australians aged 40 years and over, the weighted prevalence of vision loss from ocular trauma was 0.79% (95% CI: 0.56, 1.13), consisting of 3 participants with monocular visual impairment, 8 participants with monocular blindness and 1 participant with binocular blindness due to trauma. Males displayed a higher prevalence of vision loss from ocular trauma in the non-Indigenous (1.25% (95% CI: 0.58, 2.70) vs. 0.47% (95% CI: 0.10, 2.10), p = 0.03) and Indigenous populations (0.38% (95% CI: 0.16, 0.94) vs. 0.12% (95% CI: 0.02, 0.59), p = 0.02). Extrapolating these findings to the Australian population, we estimate that approximately 14,395 non-Indigenous Australians aged 50 years or over and 1063 Indigenous Australians aged 40 years or over have vision loss caused by ocular trauma. Among the 10 cases of monocular vision loss from ocular trauma in the non-Indigenous population, 3 eyes (30%) had traumatic cataract, 3 (30%) eyes were enucleated due to previous trauma, 2 (20%) had traumatic corneal scars and 2 (20%) had traumatic retinal detachment or scarring. Trauma resulting in optic atrophy was attributed as an underlying cause of bilateral blindness in one Indigenous participant. Of the remaining 11 cases of monocular vision loss from ocular trauma in the Indigenous population, 3 eyes (27.3%) had traumatic cataract, 3 (27.3%) eyes were enucleated due to previous trauma, 2 (18.2%) had traumatic corneal scars and 2 (18.2%) had traumatic retinal detachment or scarring. Due to the low frequency of ocular trauma causing vision loss found in the NEHS, Indigenous and non-Indigenous data were combined in Poisson regression analysis examining associations between ocular trauma causing vision loss and selected
Table 2 Relationship between ocular injury causing vision loss and selected characteristics in Australians (n = 4836). Associated factors
Indigenous Age (per 10 years) Gender (male) Education Remoteness Major city Inner Regional Outer Regional Remote Very Remote
Univariate logistic regression
Multivariable logistic regression
IRR [95% (CI)]
p
IRR [95% (CI)]
p*
3.26 (1.37, 7.75) 0.18 (0.84, 1.20) 1.14 (0.58, 2.88) 0.05 (0.02, 0.13)
0.01 0.73 0.19 0.18
3.26 (0.61, 17.41) 1.33 (0.43, 4.08) 2.96 (0.48, 18.33) 1.07 (0.97, 1.18)
0.16 0.61 0.23 0.17
1 0.63 (0.07, 5.75) 2.08 (0.39, 10.97) 2.16 (0.51, 9.03) 4.60 (1.54, 13.78)
0.68 0.38 0.28 0.01
1 0.61 (0.07, 5.09) 2.09 (0.38, 11.58) 2.24 (0.55, 9.08) 3.44 (1.24, 9.54)
0.63 0.39 0.25 0.02
IRR = incidence rate ratio; CI = Confidence Interval.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020
G Model JINJ 7421 No. of Pages 4
4
S. Keel et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Rather our data is limited to reporting the prevalence of vision loss due to ocular trauma (based on relevant clinical data). It is also possible that survivor bias could have led to an underestimation in the true prevalence of vision loss due to ocular trauma in this study, where people who experienced severe ocular or head injuries during their lifetime may have been at a higher risk of earlier mortality. In summary, the NEHS provides Indigenous-specific prevalence estimates of vision loss due to ocular trauma in a national population-based sample of Australian adults. The majority of vision loss from trauma is preventable and it impacts a significant number of Australians in their productive years. Our data may indicate that males, Indigenous Australians and those residing in Very Remote communities may benefit from improved health promotion efforts focussing on injury prevention, including the use of protective eyewear. Further research on the annual incidence, causes and location of occurrence of any ocular trauma in Australia would be helpful to design these targeted programs. Conflicts of interest Nil to declare. Acknowledgements The Centre for Eye Research Australia (CERA) and Vision 2020 Australia wish to recognise the contributions of all the NEHS project steering committee members (Professor Hugh Taylor, Dr Peter van Wijngaarden, Jennifer Gersbeck, Dr Jason Agostino, Anna Morse, Sharon Bentley, Robyn Weinberg, Christine Black, Genevieve Quilty, Louis Young and Rhonda Stilling) and the core CERA research team who assisted with the survey field work (Joshua Foreman, Pei Ying Lee, Rosamond Gilden, Larissa Andersen, Benny Phanthakesone, Celestina Pham, Alison Schokman, Megan Jackson, Hiba Wehbe, John Komser and Cayley Bush). Furthermore, we would like to acknowledge the overwhelming support from all collaborating Indigenous organisations who assisted with the implementation of the survey, and the Indigenous health workers and volunteers in each survey site who contributed to the field work. The National Eye Health Survey was funded by the Department of Health of the Australian Government, and also received financial contributions from Novartis Australia and the Peggy and Leslie Cranbourne Foundation. In-kind support was received from our industry and sector partners, OPSM, Carl Zeiss, Designs for Vision, the Royal Flying Doctor Service, Optometry Australia and the Brien Holden Vision Institute. We would like to specifically acknowledge OPSM, who kindly donated sunglasses valued at $130 for each study participant. The Centre for Eye Research Australia receives Operational Infrastructure Support from the Victorian Government. The Principal Investigator, Dr Mohamed Dirani, is supported by a NHMRC Career Development Fellowship (#1090466). The PhD
student, Joshua Foreman is supported by an Australian Postgraduate Award scholarship. References [1] Negrel AD, Thylefors B. The global impact of eye injuries. Ophthalmic Epidemiol 1998;5(3):143–69. [2] Australian Institue of Health and Welfare. Eye related injuries in Australia. 2009. . Accessed 3rd of March 2017 http://www.aihw.gov.au/WorkArea/ DownloadAsset.aspx?id=6442458828. [3] Fong LP. Eye injuries in victoria: Australia. Med J Aust 1995;162(2):64–8. [4] McCarty CA, Fu CL, Taylor HR. Epidemiology of ocular trauma in Australia. Ophthalmology 1999;106(9):1847–52. [5] Australian Institue of Health and Welfare. Australian hospital statistics 2004– 05. Canberra, Australia. Accessed 27th March 2017. 2016. http://www.aihw. gov.au/WorkArea/DownloadAsset.aspx?id=6442456820:2006. [6] Northey LC, Bhardwaj G, Curran S, McGirr J. Eye trauma epidemiology in regional Australia. Ophthalmic Epidemiol 2014;21(4):237–46. [7] Thompson CG, Griffits RK, Nardi W, Tester MP, Noble MJ, Cottee L, et al. Penetrating eye injuries in rural new south wales. Aust N Z J Ophthalmol 1997;25(1):37–41. [8] Casson RJ, Walker JC, Newland HS. Four-year review of open eye injuries at the Royal Adelaide Hospital. Clin Exp Ophthalmol 2002;30(1):15–8. [9] Smith AR, O'Hagan SB, Gole GA. Epidemiology of open- and closed-globe trauma presenting to Cairns Base Hospital, Queensland. Clin Exp Ophthalmol 2006;34(3):252–9. [10] Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia: the Blue Mountains Eye Study. Ophthalmology 1996;103(3):357–64. [11] Taylor HR, Xie J, Fox S, Dunn RA, Arnold AL, Keeffe JE. The prevalence and causes of vision loss in indigenous australians: the national indigenous eye health survey. Med J Aust 2010;192(6):312–8. [12] Landers J, Henderson T, Craig J. The prevalence and causes of visual impairment in indigenous Australians within central Australia: the Central Australian Ocular Health Study. Br J Ophthalmol 2010;94(9):1140–4. [13] Foreman J, Keel S, Dunn R, van Wijngaarden P, Taylor HR, Dirani M. Sampling methodology and site selection in the National Eye Health Survey (NEHS): an Australian population-based prevalence study. Clin Exp Ophthalmol 2016. [14] Australian Bureau of Statistics. Australian statistical geography standard (ASGS). 2014. . Accessed 2nd April 2017 http://www.abs.gov.au/websitedbs/ D3310114. nsf/home/Australian+Statistical+Geography+Standard+(ASGS). [15] Australian Institute of Health and Welfare. Vision problems among older Australians. 2005. . Accessed 10th February 2017 http://www.aihw.gov.au/ WorkArea/DownloadAsset.aspx?id=6442453390. [16] Australian Institute of Health and Welfare. Mortality and life expectancy of indigenous Australians. 2014. . Accesed 10th of February 2017 http://www. aihw.gov.au/WorkArea/DownloadAsset.aspx?id=60129548468. [17] Foreman J, Keel S, van Wijngaarden P, Taylor HR, Recruitment Dirani M. Testing protocol in the national eye health survey: a population-based eye study in Australia. Ophthalmic Epidemiol 2017;1–11. [18] Happich M, Reitberger U, Breitscheidel L, Ulbig M, Watkins J. The economic burden of diabetic retinopathy in Germany in 2002. Graefes Arch Clin Exp Ophthalmol 2008;246(1):151–9. [19] Ferris F, Wilkinson C, Bird A, Chakravarthy U, Chew E, Csaky K, et al. Clinical classification of age-related macular degeneration. Ophthalmol 2013;120:844–51. [20] Mukesh B, McCarty C, Rait J, Taylor H. Five-year incidence of open-angle glaucoma. Ophthalmol 2002;109:1047–51. [21] Berecki-Gisolf J, Smith PM, Collie A, McClure RJ. Gender differences in occupational injury incidence. Am J Ind Med 2015;58(3):299–307. [22] Australian Institute of Health and Welfare. Health behaviours and other risks to health. 2014. . Accessed 21st April 2017 http://wwwaihwgovau/australiashealth/2014/health-behaviours. [23] Arnold AL, Busija L, Keeffe JE, Taylor HR. Use of eye care services by indigenous Australian adults. Med J Aust 2011;194(10):537–8. [24] Kiely PM, Chakman J. Optometric practice in australian standard geographical classification-remoteness areas in Australia, 2010. Clin Exp Optom 2011;94 (5):468–77.
Please cite this article in press as: S. Keel, et al., The prevalence of vision loss due to ocular trauma in the Australian National Eye Health Survey, Injury (2017), http://dx.doi.org/10.1016/j.injury.2017.09.020