Ophthalmic Patient Data Registries: Defining and Improving Quality and Outcomes
2014; Elsevier BV; Volume: 121; Issue: 3 Linguagem: Inglês
10.1016/j.ophtha.2013.12.033
ISSN1549-4713
AutoresSahar Kohanim, Paul W. Sternberg,
Tópico(s)Retinal Diseases and Treatments
ResumoTwo articles in this issue of Ophthalmology (see articles on pages 637 and 643) highlight the importance of registries and the wealth of data that they can provide. In the first, Jackson et al1Jackson T.L. Donachie P.H. Sallam A. et al.United Kingdom National Ophthalmology Database Study of Vitreoretinal Surgery: report 3, retinal detachment.Ophthalmology. 2014; 121: 643-648Abstract Full Text Full Text PDF PubMed Scopus (94) Google Scholar used the United Kingdom National Ophthalmology Database to study real-life surgeon procedure choices, complication rates, and reoperation rates for primary retinal detachment repair and suggest that this information may assist vitreoretinal surgeons in benchmarking their own rates. In the second article, Bertelsen et al2Bertelsen M. Linneberg A. Christoffersen N. et al.Mortality in patients with central retinal vein occlusion.Ophthalmology. 2014; 121: 637-642Google Scholar from Denmark cross-referenced multiple Danish national registries to perform a registry-based case-control study assessing mortality and comorbidities in patients with and without central retinal vein occlusion. They found that, although mortality was increased in patients with central retinal vein occlusion, this could be accounted for mostly by the patients' pre-existing comorbidities. These studies could not be more timely. They highlight the trove of data that registries can provide researchers and how registries can be used in ways not foreseen initially. As the American Academy of Ophthalmology begins to roll out the Intelligent Research in Sight (IRIS) national ophthalmology registry, much can be learned from the experience of other countries with national ophthalmology databases and from the American experience with databases in other medical fields. We are living through an information revolution, and it is now more important than ever to create and use medical registries for improving outcomes and performing health services research. This "big data" age represents an exciting and challenging time in medicine and technology. Our capacity to store data has increased 50 million-fold over the past half century, whereas the cost of digital storage has decreased by approximately 30% per year.3Mayer-Schönberger V. Cukier K. Big data: a revolution that will transform how we live, work, and think. Houghton Mifflin Harcourt, Boston2013Google Scholar Thanks to these computational advances, vast amounts of data can now be processed and analyzed quickly. Big data has transformed many areas of our lives, such as online shopping, targeted advertisement on social media, business practices, and public health surveillance programs, and has the potential to revolutionize health care. Data are categorized as a nonrivalrous good, meaning that one person's use of it does not diminish other's ability to use it.4Mankiw N.G. Principles of microeconomics.6th ed. South-Western Cengage Learning, Mason, OH2012Google Scholar When we collect data, we often have a primary question in mind. However, there can be unforeseen secondary uses. The value of data therefore lies not only in its primary intended uses, but also in the many ways in which it potentially can be reused in the future. This makes it much more valuable over time.3Mayer-Schönberger V. Cukier K. Big data: a revolution that will transform how we live, work, and think. Houghton Mifflin Harcourt, Boston2013Google Scholar Implementing medical registries will have certain specific foreseeable benefits, such as assessing outcomes and accepted practice patterns, and thus improving patient care. The Centers for Medicare and Medicaid Services (CMS) already track procedures and diagnoses, and the United States Food and Drug Administration already uses registries for postmarket surveillance of adverse events. The United States Census also contains a wealth of epidemiologic data. Furthermore, we anticipate that reimbursement increasingly will become tied to the ability to demonstrate performance relative to specific benchmarks. That is to say, whether (and how much) insurers will pay providers for services rendered could depend on the ability of the provider to demonstrate that the level (or expense) of care that they provide is commensurate with the norm in their profession. Although these data can come from a variety of sources, registries provide a straightforward, validated, and transparent way to do so. In other fields in medicine, robust patient registries already have helped to improve patient care. Cardiology and cardiac surgery were the pioneer fields in which registries have been used to track and improve patient outcomes. Registry-based cardiology studies have demonstrated that certain invasive techniques do not actually improve patient mortality.5Frobert O. Lagerqvist B. Olivecrona G.K. et al.Thrombus aspiration during ST-segment elevation myocardial infarction.N Engl J Med. 2013; 369: 1587-1597Crossref PubMed Scopus (778) Google Scholar, 6Lauer M.S. D'Agostino Sr., R.B. The randomized registry trial—the next disruptive technology in clinical research?.N Engl J Med. 2013; 369: 1579-1581Crossref PubMed Scopus (438) Google Scholar Large cardiology patient databases in Europe have been used to assess in which situations accepted practice guidelines for the in-hospital management of heart failure are and are not followed.7Maggioni A.P. Anker S.D. Dahlstrom U. et al.Are hospitalized or ambulatory patients with heart failure treated in accordance with European Society of Cardiology guidelines? Evidence from 12 440 patients of the ESC Heart Failure Long-Term Registry.Eur J Heart Failure. 2013; 15: 1173-1184Crossref PubMed Scopus (471) Google Scholar These observations have led to suggestions that certain guidelines be changed. A registry-based study in the United States demonstrated that, apart from amiodarone, the choice of preoperative cardiac medications before coronary artery bypass grafting had no impact on the rate of postoperative atrial fibrillation.8Piccini J.P. Zhao Y. Steinberg B.A. et al.Comparative effectiveness of pharmacotherapies for prevention of atrial fibrillation following coronary artery bypass surgery.Am J Cardiol. 2013; 112: 954-960Abstract Full Text Full Text PDF PubMed Scopus (15) Google Scholar On the contrary, it has been found that the relatively common practice of giving β-blockers to patients undergoing percutaneous coronary interventions (such as stent placement) actually increased the rate of cardiac death or myocardial infarction in certain groups of patients.9Ozasa N. Morimoto T. Bao B. et al.Beta-blocker use in patients after percutaneous coronary interventions: one size fits all? Worse outcomes in patients without myocardial infarction or heart failure.Int J Cardiol. 2013; 168: 774-779Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar Registry-based studies have been used to create models that predict patients at risk for poor outcomes or death10Brown J.R. Conley S.M. Niles 2nd, N.W. Predicting readmission or death after acute ST-elevation myocardial infarction.Clin Cardiol. 2013; 36: 570-575PubMed Google Scholar and to demonstrate that predictive scoring systems outperform clinical impression and improve patient outcomes.11Bajaj R.R. Goodman S.G. Yan R.T. et al.Treatment and outcomes of patients with suspected acute coronary syndromes in relation to initial diagnostic impressions (insights from the Canadian Global Registry of Acute Coronary Events [GRACE] and Canadian Registry of Acute Coronary Events [CANRACE]).Am J Cardiol. 2013; 111: 202-207Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar On a health system-wide level, they have demonstrated that patients can be transitioned from specialized heart failure centers to their own primary care physicians without decreased ideal pharmacotherapy adherence.12Gjesing A. Schou M. Torp-Pedersen C. et al.Patient adherence to evidence-based pharmacotherapy in systolic heart failure and the transition of follow-up from specialized heart failure outpatient clinics to primary care.Eur J Heart Failure. 2013; 15: 671-678Crossref PubMed Scopus (36) Google Scholar Lastly, these types of studies can be used to assess whether prior race-based discrepancies in the provision of evidence-based medicine are improving.13Reynolds D. Albert N.M. Curtis A.B. et al.Race and improvements in the use of guideline-recommended therapies for patients with heart failure: findings from IMPROVE HF.J Natl Med Assoc. 2012; 104: 287-298PubMed Scopus (7) Google Scholar These represent some of the ways in which the mature cardiology databases have been used to understand and monitor patient care better in the past year alone. Apart from the 2 studies published in this issue,1Jackson T.L. Donachie P.H. Sallam A. et al.United Kingdom National Ophthalmology Database Study of Vitreoretinal Surgery: report 3, retinal detachment.Ophthalmology. 2014; 121: 643-648Abstract Full Text Full Text PDF PubMed Scopus (94) Google Scholar, 2Bertelsen M. Linneberg A. Christoffersen N. et al.Mortality in patients with central retinal vein occlusion.Ophthalmology. 2014; 121: 637-642Google Scholar the field of ophthalmology has also had success in demonstrating the important role for large patient registry-based studies. For example, clinical trials that evaluate infrequent end points, such as postoperative endophthalmitis, require very large sample sizes, are thus costly,14Gower E.W. Lindsley K. Nanji A.A. et al.Perioperative antibiotics for prevention of acute endophthalmitis after cataract surgery.Cochrane Database of Systematic Reviews. 2013; 7: CD006364PubMed Google Scholar and would particularly benefit from the use of large registries. As a result, studies using the Swedish National Cataract Register, which now contains more than 1 million reported cataract surgeries,15Behndig A. Montan P. Stenevi U. et al.One million cataract surgeries: Swedish National Cataract Register 1992–2009.J Cataract Refract Surg. 2011; 37: 1539-1545Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar have demonstrated the effect of incision type and location on the rate of postoperative endophthalmitis.16Lundstrom M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery: a nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar Further studies have demonstrated the effectiveness of intracameral cefuroxime (and the lack of effect of topical antibiotic drops) in reducing endophthalmitis.16Lundstrom M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery: a nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 17Friling E. Lundstrom M. Stenevi U. Montan P. Six-year incidence of endophthalmitis after cataract surgery: Swedish national study.J Cataract Refract Surg. 2013; 39: 15-21Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar Factors that contribute to refractive surprise also have been elucidated.18Behndig A. Montan P. Stenevi U. et al.Aiming for emmetropia after cataract surgery: Swedish National Cataract Register study.J Cataract Refract Surg. 2012; 38: 1181-1186Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar Registries also may be beneficial for rare diseases, where they can be studied better by aggregating the experience of multiple centers, or even entire countries. For example, central nervous system metastases from uveal melanoma are exceedingly rare, but could be quantified and studied using the Danish Cancer Registry and cross-listing these records to the national register of death certificates.19Holfort S.K. Lindegaard J. Isager P. et al.CNS metastasis from malignant uveal melanoma: a clinical and histopathological characterisation.Br J Ophthalmol. 2009; 93: 641-644Crossref PubMed Scopus (7) Google Scholar Registries also have demonstrated some interesting associations between ophthalmic conditions or procedures and other medical conditions.20Hanhart J. Vinker S. Nemet A. et al.Prevalence of epilepsy among cataract patients.Curr Eye Res. 2010; 35: 487-491Crossref PubMed Scopus (8) Google Scholar, 21Kaiserman I. Kaiserman N. Nakar S. Vinker S. Herpetic eye disease in diabetic patients.Ophthalmology. 2005; 112: 2184-2188Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 22Nemet A.Y. Vinker S. Levartovsky S. Kaiserman I. Is cataract associated with cardiovascular morbidity?.Eye (Lond). 2010; 24: 1352-1358Crossref PubMed Scopus (33) Google Scholar On the other hand, large medical and population registries present certain challenges. They can be costly to organize and maintain. However, with ophthalmologists facing penalties (in the form of reduced reimbursement) from CMS if they do not submit quality measures, any fee that the Academy of Ophthalmology charges to the member physicians should be countered by the savings they will generate (by avoiding CMS nonreporter penalties) by having IRIS compile and report their performance. Further, some concerns have been raised about patient confidentiality. Although modern registries, including the upcoming American Academy of Ophthalmology-sponsored IRIS registry, can interface directly with many electronic health record systems, physicians who participate not only must use an electronic health record, but also must have an interfacing electronic health record system. Lastly, there are challenges in independently confirming the veracity of what is entered, and the validity of the conclusions is only as good as the validity of the inputs. As the old adage goes: "garbage in, garbage out." Many of these anxieties will be addressed by IRIS, but it is likely that there will be unforeseen challenges and difficulties. Balanced against these potential negatives are the many positive applications and implications of establishing large registries: they can be used to assess the comparative clinical effectiveness of different interventions; they can facilitate health services research; they can improve provider performance through benchmarking and by allowing physicians to compare their practice patterns and their outcomes, be it to a national benchmark or to their colleagues; and they can improve accountability to the government and to the public. More important, however, are the many unforeseen ways in which collected data can be used in the future. In particular, this happens when multiple databases can be cross-referenced, as was seen in the Danish study by Bertelsen et al.2Bertelsen M. Linneberg A. Christoffersen N. et al.Mortality in patients with central retinal vein occlusion.Ophthalmology. 2014; 121: 637-642Google Scholar Similarly, in Denmark, the national cancer database has been cross-referenced with cell phone data to assess the effect of cell phone usage on the development of cancer.23Frei P. Poulsen A.H. Johansen C. et al.Use of mobile phones and risk of brain tumours: update of Danish cohort study.BMJ. 2011; 343: d6387Crossref PubMed Scopus (142) Google Scholar It is in this regard that the United States falls behind its European counterparts. This is a pity, given how much epidemiologic data can be found in the United States Census. However, largely because of privacy concerns, United States patient databases do not record identifiers (e.g., Social Security numbers) that are unique to an individual and that are carried with them throughout life and between registries. Whether or not you are comfortable with the risks of big data, the age of patient registries and databases is upon us. This should be seen as an opportunity that can benefit clinicians, researchers, patients, and the healthcare system as a whole. The articles in this journal are just a taste of what can be accomplished. However, given the voluntary nature of registry participation, we can achieve all of the promised benefits of the IRIS registry only if ophthalmologists in both private practice and in the academic sector embrace this opportunity. The uses of these data that we are able to imagine today represent only a minute fraction of the potential uses in the future. Like an iceberg, much of its weight is hidden beneath the surface. Appreciating this full value can lead to immense benefits for all involved. Mortality in Patients with Central Retinal Vein OcclusionOphthalmologyVol. 121Issue 3PreviewTo assess mortality in patients with central retinal vein occlusion (CRVO). Full-Text PDF United Kingdom National Ophthalmology Database Study of Vitreoretinal Surgery: Report 3, Retinal DetachmentOphthalmologyVol. 121Issue 3PreviewTo describe rhegmatogenous retinal detachment (RD) surgery. Full-Text PDF
Referência(s)