A scoping review of studies using observational data to optimise dynamic treatment regimens

BACKGROUND

Dynamic treatment regimens (DTRs) formalise the multi-stage and dynamic decision problems that clinicians often face when treating chronic or progressive medical conditions. Compared to randomised controlled trials, using observational data to optimise DTRs may allow a wider range of treatments to be evaluated at a lower cost. This review aimed to provide an overview of how DTRs are optimised with observational data in practice.

METHODS

Using the PubMed database, a scoping review of studies in which DTRs were optimised using observational data was performed in October 2020. Data extracted from eligible articles included target medical condition, source and type of data, statistical methods, and translational relevance of the included studies.

RESULTS

From 209 PubMed abstracts, 37 full-text articles were identified, and a further 26 were screened from the reference lists, totalling 63 articles for inclusion in a narrative data synthesis. Observational DTR models are a recent development and their application has been concentrated in a few medical areas, primarily HIV/AIDS (27, 43%), followed by cancer (8, 13%), and diabetes (6, 10%). There was substantial variation in the scope, intent, complexity, and quality between the included studies. Statistical methods that were used included inverse-probability weighting (26, 41%), the parametric G-formula (16, 25%), Q-learning (10, 16%), G-estimation (4, 6%), targeted maximum likelihood/minimum loss-based estimation (4, 6%), regret regression (3, 5%), and other less common approaches (10, 16%). Notably, studies that were primarily intended to address real-world clinical questions (18, 29%) tended to use inverse-probability weighting and the parametric G-formula, relatively well-established methods, along with a large amount of data. Studies focused on methodological developments (45, 71%) tended to be more complicated and included a demonstrative real-world application only.

CONCLUSIONS

As chronic and progressive conditions become more common, the need will grow for personalised treatments and methods to estimate the effects of DTRs. Observational DTR studies will be necessary, but so far their use to inform clinical practice has been limited. Focusing on simple DTRs, collecting large and rich clinical datasets, and fostering tight partnerships between content experts and data analysts may result in more clinically relevant observational DTR studies.

Prevalence and associations of non-retinopathy ocular conditions among older Australians with self-reported diabetes: The National Eye Health Survey

AIM

To determine the prevalence and associations of non-retinopathy ocular conditions among older Australian adults with diabetes.

METHODS

Multistage random-cluster sampling was used to select 3098 non-indigenous Australians aged 50y or older (46.4% male) and 1738 indigenous Australians aged 40y or older (41.1% male) from all levels of geographic remoteness in Australia. Participants underwent a standardised questionnaire to ascertain diabetes history, and a clinical examination to identify eye disease. We determined the prevalence of uncorrected refractive error, visually significant cataract, cataract surgery, age-related macular degeneration, glaucoma, ocular hypertension, retinal vein occlusion and epiretinal membrane among those with and without self-reported diabetes.

RESULTS

Participants with self-reported diabetes had a higher prevalence of cataract surgery than those without diabetes (28.8% vs 16.9%, OR 1.78, 95%CI: 1.35-2.34 among non-indigenous Australians, and 11.3% vs 5.2%, OR 1.62, 95%CI: 1.22-2.14 among indigenous Australians). Diabetic retinopathy (DR) increased the odds of cataract surgery among self-reported diabetic indigenous and non-indigenous Australians (OR 1.89, P=0.004 and OR 2.33, P<0.001 respectively). Having diabetes for ≥20y and having vision-threatening DR increased the odds of cataract surgery among indigenous Australians with diabetes (OR 3.73, P=0.001 and 7.58, P<0.001, respectively).

CONCLUSION

Most non-retinopathy ocular conditions are not associated with self-reported diabetes. However, to account for Australia's worsening diabetes epidemic, interventions to reduce the impact of diabetes-related blindness should include increased cataract surgery services.

Relationship Between Rod-Mediated Sensitivity, Low-Luminance Visual Acuity, and Night Vision Questionnaire in Age-Related Macular Degeneration

Purpose

To quantify the association between dark adaptation parameters and other clinical measures of visual function among people with and without early and intermediate age-related macular degeneration (AMD).

Methods

In this cross-sectional study, participants underwent multimodal imaging and visual function testing, including best-corrected visual acuity (BCVA), low-luminance visual acuity (LLVA), low-luminance deficit (LLD = BCVA – LLVA) and the 10-item Night Vision Questionnaire (NVQ-10). Dynamic and static dark-adapted chromatic perimetry (DACP) was performed. Sensitivity difference was defined as the difference in sensitivity between the 505-nm and 625-nm stimuli. Rod intercept time (RIT) was estimated as the time required to reach a threshold of −3 log candelas/meter² with the 505-nm stimulus following bleaching. The magnitude of association between the DACP parameters and other clinical tests was estimated via mixed-effects regression.

Results

A total of 51 participants (aged 51–88 years, 65% female, 39% with AMD) were included. RIT was found to be negatively associated with BCVA (P < 0.001), LLVA (P = 0.005), and NVQ-10 score (P = 0.028) but not LLD (P = 0.763). There was no evidence of an association between sensitivity difference and any of the clinical measures (P ≥ 0.081).

Conclusions

Reduced rod function, as determined by RIT, was associated with lower NVQ-10 scores (designed to interrogate rod-mediated function) and with worse BCVA and LLVA (measures of cone function).

Translational Relevance

Decreasing rod function maybe indicative of more generalized photoreceptor dysfunction involving cones. Further development of questionnaires to target function in scotopic conditions may provide an easier to administer test without the need to perform perimetric tests of rod function.

Gram-negative endophthalmitis: A prospective study examining the microbiology, clinical associations and visual outcomes following infection

IMPORTANCE

This is the largest Gram-negative endophthalmitis specific series and provides important evidence to guide management.

BACKGROUND

Endophthalmitis is a sight-threatening emergency. Gram-negative infections are associated with poorer visual outcomes; however, there is limited literature pertaining to this uncommon condition.

DESIGN

Prospective case series.

PARTICIPANTS

All patients presenting with endophthalmitis to a tertiary institution over a 20-year period.

METHODS

Data were collected prospectively and entered into a registry. Patients with microbiological evidence of Gram-negative infection were included in the analysis.

MAIN OUTCOMES MEASURES

Final visual acuity (VA), precipitating events, causative organisms, antibiotic sensitivity profiles and risk factors for poor visual outcomes were reported.

RESULTS

One hundred Gram-negative organisms were isolated in 97 eyes. Final VA was worse than 6/60 in 65 (67.0%) eyes at follow-up and 29 (29.9%) eyes were eviscerated or enucleated. Microbial keratitis (26.8%, n = 26) and Pseudomonas aeruginosa (34.0%, n = 34) were the most common precipitating event and causative organism, respectively. Eight (8.0%) isolates were third-generation cephalosporin resistant; of which, 7 (88.0%) were sensitive to ciprofloxacin. Preceding microbial keratitis (OR = 13.16, P = .015) or P. aeruginosa infections (OR = 3.40, P = .045) were strongly associated with poorer visual outcomes (worse than 6/60).

CONCLUSIONS AND RELEVANCE

Visual outcomes following Gram-negative endophthalmitis are extremely poor, with almost 30% of patients being eviscerated or enucleated. A majority of ceftazidime resistant organisms are sensitive to ciprofloxacin, providing evidence to support the empirical use of quinolones. Clinicians should be mindful that infections secondary to P. aeruginosa or microbial keratitis carry a particularly poor prognosis.

Association between Patient-Reported Outcomes and Time to Late Age-Related Macular Degeneration in the Laser Intervention in Early Stages of Age-Related Macular Degeneration Study

PURPOSE

To investigate the relationship between patient-reported outcome (PRO) questionnaire responses and time to late age-related macular degeneration (AMD; neovascular AMD [nAMD] or multimodal imaging [MMI]-defined atrophy) among individuals with bilateral large drusen, and the prognostic value of baseline PROs for 36-month AMD status.

DESIGN

Exploratory analyses from a multicenter randomized controlled trial of an AMD intervention (Australian New Zealand Clinical Trials Registry identifier, ACTRN12612000704897).

PARTICIPANTS

Sham treatment group of the Laser Intervention in Early Stages of Age-Related Macular Degeneration (LEAD) Study (n = 141; age, 50-88 years; 77% female).

METHODS

The 28-item Impact of Vision Impairment (IVI-28) and 10-item Night Vision Questionnaire (NVQ-10) were administered at the baseline visit. The PRO scores were derived using rating scale models. Multivariate Cox regression adjusting for demographics and clinical measures of vision (low-luminance visual acuity, low-luminance deficit, and microperimetric sensitivity) from the poorer-performing eye was used to investigate the association between PRO scores and time to late AMD in either eye. Multivariate competing-risk regression was used to estimate cause-specific subhazard ratios for nAMD and atrophy in either eye. Cross-validated logistic lasso models were used to estimate the predicted probability of AMD at 36 months. The area under the receiver operating characteristic curve was assessed to compare prognostic accuracy between models with and without PROs.

MAIN OUTCOME MEASURE

Time until nAMD or atrophy in either eye.

RESULTS

The PRO scores were skewed toward higher functional vision. Higher IVI-28 scores were associated with a lower risk of progression to MMI-defined atrophy (20 events: adjusted hazard ratio, 0.65/logit increase; P = 0.002) but not nAMD (10 events; P = 0.562). Insufficient evidence was found of an association between NVQ-10 score and rate of progression to late AMD (P ≥0.149). Baseline IVI-28 scores were found to contribute to the prognosis of atrophy at the 36-month visit (P = 0.010).

CONCLUSIONS

On average, PROs were associated with an increased risk of progression from intermediate AMD to MMI-defined atrophy. Continuing development of instruments to record PROs in the early stages of AMD have the potential to produce inexpensive and efficient tools to assist in the assessment of disease severity and risk of AMD progression.

Correction to: A comparison of methods to estimate the survivor average causal effect in the presence of missing data: a simulation study

In the original publication of this article [1], the incorrect causal diagram was submitted as Fig. 1.

Quantitative Fundus Autofluorescence and Genetic Associations in Macular, Cone, and Cone-Rod Dystrophies

PURPOSE

To investigate quantitatively lipofuscin-associated fundus autofluorescence in patients with macular and cone/cone-rod dystrophies (MD/CCRDs).

DESIGN

Prospective, single-center, case-control study.

PARTICIPANTS

Two hundred thirty patients with MD/CCRDs who had undergone genetic testing and 110 control participants without any eye disease.

METHODS

Participants were examined using quantitative fundus autofluorescence (qAF) imaging with a modified confocal scanning laser ophthalmoscope equipped with an internal fluorescent reference (modified Spectralis HRA-OCT; Heidelberg Engineering, Heidelberg, Germany). Mean qAF values were obtained by averaging measurements from an 8-segment ring centered on the fovea (qAF₈) and compared with controls.

MAIN OUTCOME MEASURES

The qAF₈ levels.

RESULTS

Elevated qAF₈ values were a frequent finding (n = 105 [45%]) and associated with ABCA4 (n = 73 [70%]), PRPH2 (n = 9 [9%]), CERKL (n = 3 [3%]), PROM1 (n = 2 [2%]), CRX (n = 1 [1%]), and CDHR1 (n = 1 [1%]) mutations. Reduced qAF₈ values were rare (n = 15 [7%]) and found predominantly among patients with MERTK (n = 3 [20%]) and RDH5 (n = 2 [13%]) mutations. Patients with normal qAF₈ values (n = 110 [48%]) showed high genotypic heterogeneity. For various genes including ABCA4, PRPH2, CDHR1, and PROM1, higher qAF₈ measures were associated with specific phenotypes and genotypes. For instance, qAF₈ values were normal in PRPH2-related central areolar chorioretinal dystrophy but increased in PRPH2-related Stargardt-like retinopathy. Accordingly, high qAF₈ levels were associated with specific genetic causes and mutation detection rates in characteristic but genetically heterogenous clinical phenotypes, such as a Stargardt-like flecked fundus, bull's eye maculopathy, or pattern dystrophy. In genetically unsolved cases (16 with elevated, 35 with normal, 7 with reduced qAF values), qAF₈ was used to support or reject ambiguous results of genetic testing, to suggest underlying pathogenic pathways, and to predict disease in otherwise healthy participants.

CONCLUSIONS

Quantitative fundus autofluorescence imaging revealed characteristic qAF levels in association with certain gene mutations and in participants without detected mutations. These findings indicate that qAF may facilitate differential diagnostics of MD/CCRDs and may offer novel pathogenetic insights that may be of particular value for the assessment of future treatment approaches.

Comparative Evaluation Of Clinical Characteristics And Visual Outcomes Of Traumatic And Non-Traumatic Graft Dehiscence Following Corneal Transplantation Surgery

Purpose

To compare patient demographics, clinical associations and visual outcomes between traumatic and non-traumatic wound dehiscence, following corneal transplantation.

Methods

Retrospective review of all patients presenting with post-keratoplasty wound dehiscence to the Royal Victorian Eye and Ear Hospital between January 2005 and December 2017. Patients with wound dehiscence following keratoplasty of any cause were included.

Results

Of 71 eyes from 71 patients included, 60 (85%) were penetrating keratoplasty patients. The mean age was 56.4 years (SD=22.7, range 17.6-97) and 62% (n = 44) of patients were male. There were 28 (39%) cases of traumatic dehiscence and 43 (61%) cases of non-traumatic dehiscence. The median time interval from keratoplasty to dehiscence was significantly less in non-traumatic patients than traumatic patients (0.2 years, IQR 0.1-2.0 vs 2.3 years, IQR 0.3-14.8, p=0.01). There was no significant difference in best-corrected visual acuity at 6 months between traumatic and non-traumatic dehiscence (6/60 vs 6/36, p=0.62), suture technique (continuous vs interrupted, p=0.12), or graft type (penetrating keratoplasty vs deep anterior lamellar keratoplasty) after adjusting for keratoconus (p=0.41).

Conclusion

Post-keratoplasty wound dehiscence is a serious complication and can cause significant loss of vision. While the risk of dehiscence is lifelong, the first 3 years post-keratoplasty carry the highest risk, with non-traumatic dehiscence tending to occur earlier than traumatic dehiscence.

A comparison of methods to estimate the survivor average causal effect in the presence of missing data: a simulation study

BACKGROUND

Attrition due to death and non-attendance are common sources of bias in studies of age-related diseases. A simulation study is presented to compare two methods for estimating the survivor average causal effect (SACE) of a binary exposure (sex-specific dietary iron intake) on a binary outcome (age-related macular degeneration, AMD) in this setting.

METHODS

A dataset of 10,000 participants was simulated 1200 times under each scenario with outcome data missing dependent on measured and unmeasured covariates and survival. Scenarios differed by the magnitude and direction of effect of an unmeasured confounder on both survival and the outcome, and whether participants who died following a protective exposure would also die if they had not received the exposure (validity of the monotonicity assumption). The performance of a marginal structural model (MSM, weighting for exposure, survival and missing data) was compared to a sensitivity approach for estimating the SACE. As an illustrative example, the SACE of iron intake on AMD was estimated using data from 39,918 participants of the Melbourne Collaborative Cohort Study.

RESULTS

The MSM approach tended to underestimate the true magnitude of effect when the unmeasured confounder had opposing directions of effect on survival and the outcome. Overestimation was observed when the unmeasured confounder had the same direction of effect on survival and the outcome. Violation of the monotonicity assumption did not increase bias. The estimates were similar between the MSM approach and the sensitivity approach assessed at the sensitivity parameter of 1 (assuming no survival bias). In the illustrative example, high iron intake was found to be protective of AMD (adjusted OR 0.57, 95% CI 0.40-0.82) using complete case analysis via traditional logistic regression. The adjusted SACE odds ratio did not differ substantially from the complete case estimate, ranging from 0.54 to 0.58 for each of the SACE methods.

CONCLUSIONS

On average, MSMs with weighting for exposure, missing data and survival produced biased estimates of the SACE in the presence of an unmeasured survival-outcome confounder. The direction and magnitude of effect of unmeasured survival-outcome confounders should be considered when assessing exposure-outcome associations in the presence of attrition due to death.

Presymptomatic Retinal Sensitivity Changes in Intermediate Age-Related Macular Degeneration Associated With New Retinal Fluid

Purpose

To determine whether change in retinal sensitivity in areas with subretinal or intraretinal fluid secondary to age-related macular degeneration (AMD) precedes visual symptoms. If confirmed, retinal sensitivity testing could be used for home monitoring in AMD.

Methods

Individuals with intermediate AMD enrolled in a longitudinal study were seen every 6 months and underwent best-corrected visual acuity testing (BCVA), spectral domain–optical coherence tomography (SD-OCT), and microperimetry. Asymptomatic individuals who developed incidental, reading center–determined retinal fluid detected on SD-OCT were identified. The point-wise sensitivity (PWS) at the time of fluid detection was compared with 6 and 12 months prior.

Results

Fourteen of 161 individuals developed fluid without symptoms. PWS over fluid areas at detection was reduced compared with 6 (difference −2.04 dB, P < 0.001) and 12 months (−2.27 dB, P < 0.001) prior. PWS over fluid areas was reduced compared with perifluid areas (difference −1.02 dB, P = 0.03), peripheral areas (−1.51 dB, P < 0.001), nonprogressed fellow eyes (−1.49 dB, P = 0.006), and nonprogressed age-matched intermediate AMD eyes (−2.29 dB, P = 0.001). No difference in BCVA was observed in eyes developing fluid compared to eyes that do not develop fluid (P = 0.76).

Conclusions

Retinal areas with fluid on SD-OCT had a corresponding reduction in retinal sensitivity at the time of fluid detection compared with 6 and 12 months prior, in asymptomatic intermediate AMD without change in BCVA.

Translational Relevance

Development of self-monitoring tools to detect changes in retinal sensitivity may be helpful for early detection of retinal fluid suggestive of progression to neovascular AMD before acuity is affected.

Prevalence, associations and characteristics of severe uncorrected refractive error in the Australian National Eye Health Survey

IMPORTANCE

In Australia, nationally representative data of the burden and associations of severe uncorrected refractive error are scarce.

BACKGROUND

To report the prevalence and characteristics of severe uncorrected refractive error in Indigenous and non-Indigenous Australians.

DESIGN

Population-based cross-sectional study.

PARTICIPANTS

A total of 3098 non-Indigenous Australians aged 50 to 98 and 1738 Indigenous Australians aged 40 to 92 living in 30 randomly selected Australian sites were examined.

METHODS

Severe uncorrected refractive error was defined as an improvement of ≥2 lines on the logMAR chart in one or both eyes in participants with a presenting visual acuity <6/12.

MAIN OUTCOME MEASURE

Severe uncorrected refractive error RESULTS: Prevalence of severe uncorrected refractive error was 11.0% (95% confidence interval 9.3-13.0) in non-Indigenous and 14.5% (12.5-16.7) in Indigenous Australians. Eighty-two percent of non-Indigenous and 77% of Indigenous participants had a spherical equivalent refraction between -2.00D and +2.00D. Indigenous Australians who were older (odds ratio [OR] for 70-79 years vs 40-49 years = 3.59), resided in outer regional areas (OR = 1.78) and did not have an eye examination in the previous 2-years (OR = 1.50) were associated with higher odds of severe uncorrected refractive error. Geographical remoteness (OR = .68 for inner regional), male gender (OR = 1.30), older age (OR for 70-79 years vs 50-59 years = 1.51) and failure to have an eye examination in the previous 2-years (OR = 2.06) were associated with severe uncorrected refractive error among non-Indigenous participants.

CONCLUSIONS AND RELEVANCE

Increased public awareness of the importance of regular optometric examinations may be required in groups at high risk of severe uncorrected refractive error.

Properties of the Impact of Vision Impairment and Night Vision Questionnaires Among People With Intermediate Age-Related Macular Degeneration

Purpose

To explore the psychometric properties of the Impact of Vision Impairment (IVI-28) and Night Vision Questionnaires (NVQ-10) among people with intermediate age-related macular degeneration (iAMD).

Methods

Baseline responses were collected from 288 participants (aged 50-88 years, 74% female) in the Laser intervention in Early stages of Age-related macular Degeneration (LEAD) study in Australia and Northern Ireland. Psychometric properties (discrimination, ordering of thresholds, person separation, item miss-fit, and differential item functioning according to sex) were explored using grouped rating scale and partial credit models. Spearman's correlation was estimated to assess the association with measures of visual function (mean mesopic microperimetric sensitivity, best-corrected visual acuity, low-luminance visual acuity, and low-luminance deficit). The psychometric properties were then explored following recalibration of the instruments.

Results

In this homogenous population, ceiling effects caused by relatively high levels of functional vision were evident for both instruments. The IVI-28 and NVQ-10 displayed suboptimal discrimination between levels of functional vision in iAMD and poor targeting among people with iAMD. The correlation between ability scores and measures of visual function was mild. In general, the NVQ-10 showed superior psychometric properties to the IVI-28 among these participants. No significant improvement in reliability could be gained following recalibration.

Conclusions

Both instruments were designed for populations with more severe visual loss and poorly discriminate in this cohort of iAMD.

Translational Relevance

New instruments that can capture the subtle changes in functional vision that occur early in AMD are required to aid evaluation of emerging interventions for iAMD.

Bleb-associated endophthalmitis: Proportion, presentation, management and outcomes in Victoria, Australia

IMPORTANCE

Bleb-associated endophthalmitis is a potentially vision-threatening complication of trabeculectomy. With improvements in surgical technique and changing patterns of intraoperative antimetabolite use, a re-evaluation of the incidence of bleb-associated endophthalmitis is warranted.

BACKGROUND

To investigate changes in the incidence, presentation, management and outcomes of bleb-associated endophthalmitis between 1997 and 2015 in Victoria, Australia.

DESIGN

A retrospective cohort analysis.

PARTICIPANTS

Consecutive cases of bleb-associated endophthalmitis managed at the Royal Victorian Eye and Ear Hospital (RVEEH) between 1997 and 2015.

METHODS

Medical record review of consecutive cases of bleb-associated endophthalmitis and statistical analysis were performed.

MAIN OUTCOME MEASURES

Visual acuity, including loss of light perception, intraocular pressure, and need for further surgery.

RESULTS

Sixty-seven eyes with bleb-associated endophthalmitis (BAE) were identified. Of these, 41 had trabeculectomy performed in Victoria during the study period, over which time 11 129 trabeculectomies were performed. The proportion of BAE was stable over time (0.4%). The mean age at presentation was 73.7 ± 12.1 years old and the majority of patients were Caucasian (79.1%). The mean duration between glaucoma filtration surgery and the development of bleb-associated endophthalmitis was 3 years (Interquartile Range = 0.4-6.0 years). The cultures were positive in 71.6% of cases. Approximately 1 in 8 patients required enucleation. The final visual acuity was poor with a Snellen Visual Acuity (VA) of 6/60 or worse in two-thirds of patients.

CONCLUSIONS AND RELEVANCE

Bleb-associated endophthalmitis is an uncommon complication following glaucoma filtration surgery. The proportion has remained stable over time. Visual outcomes remain poor.

Diabetic retinopathy screening in incident diabetes mellitus type 2 in Germany between 2004 and 2013 - A prospective cohort study based on health claims data

OBJECTIVE

To assess factors associated with diabetic retinopathy (DR) screening uptake following a diagnosis of type 2 diabetes mellitus (type 2 diabetes) in Germany.

MATERIALS AND METHODS

A nationally representative prospective sample of individual-level health claims data for 250,000 members from Germany's largest public insurance provider in 2004-2013 was assessed. In the sample, 26,560 persons with incident type 2 diabetes were identified. Factors associated with subsequent DR screening were assessed using descriptive statistics, Kaplan-Meier estimator, and Cox regression analysis.

RESULTS

On average 27.6 visits to an ophthalmologist per 100 person-years in persons with incident type 2 diabetes occurred. Half of all incident cases (Kaplan-Meier estimator) had not seen an ophthalmologist after more than two years (2.25 years) following their diabetes diagnosis. In the multivariate analysis, an older age (from hazard ratio HR(70-74) = 0.93 [95%-CI: 0.89-0.97] to HR(90+) = 0.50 [95%-CI: 0.42-0.60] compared to persons aged 50-69 years) and a higher disability level (i.e. HR(disability level 3) = 0.30 [95%-CI: 0.25-0.36]) were associated with a lower likelihood, while female sex (HR = 1.12 [95%-CI: 1.08-1.15]), six or more comorbidities (HR = 1.26 [95%-CI: 1.15-1.37]), moderate (HR = 1.51 [95%-CI: 1.46-1.56]) or severe type 2 diabetes (HR = 1.53 [95%-CI: 1.45-1.61]) as well as being enrolled in a type 2 diabetes disease management program (HR = 1.78 [95%-CI: 1.69-1.87]) were associated with a higher likelihood of DR screening.

CONCLUSIONS

A high proportion of newly diagnosed persons with type 2 diabetes did not follow current German recommendations for DR screening, impeding timely detection and management of potential complications. This was more apparent among persons who were men, older or had a disability. The uptake of screening was considerably greater among those enrolled in a diseases management program. These factors need to be considered when planning DR screening services and/or referrals.

Relationship between reticular pseudodrusen and choroidal thickness in intermediate age-related macular degeneration

IMPORTANCE

Reticular pseudodrusen (RPD) is strongly associated with late age-related macular degeneration (AMD) but their aetiology remains unknown. RPD have been associated with reduced choroidal thickness (ChT) but most studies are limited by small sample size and varying severity of AMD.

BACKGROUND

To investigate the relationship between choroidal thickness and RPD in eyes with intermediate AMD (iAMD), controlling for variables known to influence ChT.

DESIGN

Retrospective cohort study.

PARTICIPANTS

Participants were recruited from Centre for Eye Research Australia.

METHODS

Colour fundus photographs, fundus auto fluorescence, near-infrared and spectral-domain ocular coherence tomography (OCT) were graded for RPD. ChT was measured from enhanced-depth imaging OCT scans at the centre of fovea, 1500 and 3000 μm nasal, temporal, superior and inferior from centre of fovea.

MAIN OUTCOME MEASURES

ChT between RPD and non-RPD group.

RESULTS

A total of 297 eyes from 152 subjects were included. A total of 84 (28%) had RPD and were older than non-RPD group (75.1 ± 5.4 years and 68.7 ± 6.9 years, respectively; P < 0.001). In unadjusted analysis, the RPD group was significantly associated with thinner choroids across all measured locations (P ≤ 0.022). After adjustment for variables, the presence of RPD was no longer associated with ChT (P ≥ 0.132 for all locations) but age (P < 0.001) and refractive error (P = 0.002) remained significantly associated with ChT.

CONCLUSIONS AND RELEVANCE

Age and refractive error, rather than RPD, was significantly associated with reduced ChT in eyes with iAMD. Choroidal insufficiency may be a less important variable in RPD aetiology than previously considered.

Physical Activity and Age-related Macular Degeneration: A Systematic Literature Review and Meta-analysis

PURPOSE

To better understand the association, in a white population, of physical activity and age-related macular degeneration (AMD)-the main cause of irreversible severe vision loss in developed countries-given the suggestion that a healthy lifestyle may assist in delaying the onset and progression of AMD.

DESIGN

Systematic review and meta-analysis.

METHODS

Medline, EMBASE, and Google Scholar were systematically searched for studies up to May 2015. Reference lists of published articles were hand searched and study authors were contacted to provide additional data. Those in the lowest category of activity in each study were compared with all other participants to assess the association between physical activity and both early and late AMD using random-effects meta-analysis.

RESULTS

Nine studies (subject age range 30-97 years) were included in the meta-analysis. Physical activity was found to have a protective association with both early AMD (8 studies, n = 38 112, odds ratio (OR) 0.92, 95% confidence interval [CI] 0.86-0.98) and late AMD (7 studies, n = 28 854, OR 0.59, 95% CI 0.49-0.72).

CONCLUSIONS

Physical activity is associated with lower odds of early and late AMD in white populations. These findings have important implications, reinforcing the public health message of staying active throughout life. However, further longitudinal studies are required to confirm and further characterize a protective effect of physical activity on the onset and/or progression of AMD.

Survival Bias When Assessing Risk Factors for Age-Related Macular Degeneration: A Tutorial with Application to the Exposure of Smoking

PURPOSE

We illustrate the effect of survival bias when investigating risk factors for eye disease in elderly populations for whom death is a competing risk. Our investigation focuses on the relationship between smoking and late age-related macular degeneration (AMD) in an observational study impacted by censoring due to death.

METHODS

Statistical methodology to calculate the survivor average causal effect (SACE) as a sensitivity analysis is described, including example statistical computing code for Stata and R. To demonstrate this method, we examine the causal effect of smoking history at baseline (1990-1994) on the presence of late AMD at the third study wave (2003-2007) using data from the Melbourne Collaborative Cohort Study.

RESULTS

Of the 40,506 participants eligible for inclusion, 38,092 (94%) survived until the start of the third study wave, 20,752 (51%) were graded for AMD (60% female, aged 47-85 years, mean 65 ± 8.7 years). Late AMD was detected in 122 participants. Logistic regression showed strong evidence of an increased risk of late AMD for current smokers compared to non-smokers (adjusted naïve odds ratio 2.99, 95% confidence interval, CI, 1.74-5.13). Among participants expected to be alive at the start of follow-up regardless of their smoking status, the estimated SACE odds ratio comparing current smokers to non-smokers was at least 3.42 (95% CI 1.57-5.15).

CONCLUSIONS

Survival bias can attenuate associations between harmful exposures and diseases of aging. Estimation of the SACE using a sensitivity analysis approach should be considered when conducting epidemiological research within elderly populations.