Institute of Applied Health Research, University of Birmingham, Birmingham, UK
Institute of Applied Health Research, University of Birmingham, Birmingham, UK.
BMJ Open. 2023 Apr 3;13(4):e073015. doi: 10.1136/bmjopen-2023-073015.
The number of people with diabetes mellitus is increasing globally and consequently so too is diabetic retinopathy (DR). Most patients with diabetes are monitored through the diabetic eye screening programme (DESP) until they have signs of retinopathy and these changes progress, requiring referral into hospital eye services (HES). Here, they continue to be monitored until they require treatment. Due to current pressures on HES, delays can occur, leading to harm. There is a need to triage patients based on their individual risk. At present, patients are stratified according to retinopathy stage alone, yet other risk factors like glycated haemoglobin (HbA1c) may be useful. Therefore, a prediction model that combines multiple prognostic factors to predict progression will be useful for triage in this setting to improve care.We previously developed a Diabetic Retinopathy Progression model to Treatment or Vision Loss (DRPTVL-UK) using a large primary care database. The aim of the present study is to externally validate the DRPTVL-UK model in a secondary care setting, specifically in a population under care by HES. This study will also provide an opportunity to update the model by considering additional predictors not previously available.
We will use a retrospective cohort of 2400 patients with diabetes aged 12 years and over, referred from DESP to the NHS hospital trusts with referable DR between 2013 and 2016, with follow-up information recorded until December 2021.We will evaluate the external validity of the DRPTVL-UK model using measures of discrimination, calibration and net benefit. In addition, consensus meetings will be held to agree on acceptable risk thresholds for triage within the HES system.
This study was approved by REC (ref 22/SC/0425, 05/12/2022, Hampshire A Research Ethics Committee). The results of the study will be published in a peer-reviewed journal, presented at clinical conferences.
ISRCTN 10956293.
全球糖尿病患者人数正在增加，因此糖尿病性视网膜病变（DR）的患者也在增加。大多数糖尿病患者通过糖尿病眼病筛查计划（DESP）进行监测，直到出现视网膜病变迹象且这些变化进展，需要转介到医院眼科服务（HES）。在那里，他们继续接受监测，直到需要治疗。由于 HES 目前面临压力，可能会出现延迟，从而导致伤害。需要根据患者的个体风险进行分诊。目前，患者仅根据视网膜病变阶段进行分层，但其他风险因素，如糖化血红蛋白（HbA1c）可能也有用。因此，结合多个预后因素来预测进展的预测模型将有助于在这种情况下进行分诊，以改善护理。我们之前使用大型初级保健数据库开发了一种用于治疗或视力丧失的糖尿病性视网膜病变进展模型（DRPTVL-UK）。本研究的目的是在二级保健环境中（特别是在 HES 护理下的人群中）对 DRPTVL-UK 模型进行外部验证。本研究还将提供一个机会，通过考虑以前不可用的其他预测因素来更新该模型。
我们将使用一个回顾性队列，纳入 2400 名年龄在 12 岁及以上、因 DESP 转诊至 NHS 医院信托的糖尿病患者，这些患者在 2013 年至 2016 年间因可转诊的 DR 就诊，随访信息记录至 2021 年 12 月。我们将使用区分度、校准和净收益等指标来评估 DRPTVL-UK 模型的外部有效性。此外，还将举行共识会议，以商定 HES 系统内分诊的可接受风险阈值。
该研究得到了 REC（参考 22/SC/0425，2022 年 12 月 5 日，汉普郡 A 研究伦理委员会）的批准。研究结果将发表在同行评议的期刊上，并在临床会议上展示。
ISRCTN 10956293。
