Добавил:
kiopkiopkiop18@yandex.ru t.me/Prokururor I Вовсе не секретарь, но почту проверяю Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_2767_Библиотеки_им_академика_М_И_Перельмана

.pdf
Скачиваний:
0
Добавлен:
31.08.2026
Размер:
25 Мб
Скачать
9.2.3 Linked SROC plots 210
https://t.me/medicina_free
9.2.3.1 Example 1: Anti- CCP forthe diagnosis ofrheumatoid arthritis– descriptive plots 210
9.2.4 Tables ofresults 211
9.3 Meta- analytical summaries 211
9.3.1 Should Iestimate anSROC curve or asummary point? 212
9.3.2 Heterogeneity 214
9.4 Fitting hierarchical models 215
9.4.1 Bivariate model 216
9.4.2 Example 1 continued: anti- CCP forthe diagnosis ofrheumatoid arthritis 217
The Rutter andGatsonis HSROC model 219
9.4.3
9.4.4 Example 2: Rheumatoid factor asa marker forrheumatoid arthritis 220
9.4.5 Data reported at multiple thresholds per study 221
9.4.6 Investigating heterogeneity 222
9.4.6.1 Criteria formodel selection 223
9.4.6.2 Heterogeneity andregression analysis using thebivariate
model 223
9.4.6.3 Example 1 continued: Investigation ofheterogeneity in
diagnostic performance ofanti- CCP 224
9.4.6.4 Heterogeneity andregression analysis using theRutter
andGatsonis HSROC model 227
9.4.6.5 Example 2 continued: Investigating heterogeneity indiagnostic
accuracy ofrheumatoid factor(RF) 228
9.4.7 Comparing index tests 230
9.4.7.1 Test comparisons based onall available studies 230
9.4.7.2 Test comparisons using thebivariate model 231
9.4.7.3 Example 3: CT versus MRI forthe diagnosis ofcoronary artery
disease 232
9.4.7.4 Test comparisons using theRutter andGatsonis HSROC
model 234
9.4.7.5
Test comparison based onstudies that directly compare
tests 235
9.4.7.6 Example 3 continued: CT versus MRI forthe diagnosis of
coronary artery disease 236
9.4.8 Approaches toanalysis withsmall numbers ofstudies 238
9.4.9 Sensitivity analysis 239
9.5 Special topics 241
9.5.1 Imperfect reference standard 241
9.5.2 Investigating andhandling verification bias 241
9.5.3 Investigating andhandling publication bias 242
9.5.4 Developments inmeta- analysis forsystematic reviews oftest accuracy 243
9.6 Chapter information 243
9.7 References 244
Contents
xi
Contents
https://t.me/medicina_free
10 Undertaking meta- analysis 249
10.1 Introduction 249
10.2 Estimation ofa summary point 251
10.2.1 Fitting the bivariate model using SAS 251
10.2.2 Fitting thebivariate model using Stata 253
10.2.3 Fitting thebivariate model using R 256
10.2.4 Bayesian estimation ofthe bivariate model 261
10.2.4.1 Specification ofthe bivariate model inrjags 261
10.2.4.2 Monitoring convergence 263
10.2.4.3 Summary statistics 264
10.2.4.4 Generating an SROC plot 265
10.2.4.5
Sensitivity analyses 266
10.3 Estimation ofa summary curve 266
10.3.1 Fitting the HSROC model using SAS 268
10.3.2 Bayesian estimation of the HSROC model 268
10.3.2.1 Specification of the HSROC model in rjags 268
10.3.2.2 Monitoring convergence 270
10.3.2.3 Summary statistics and SROC plot 271
10.3.2.4 Sensitivity analyses 272
10.4 Comparison ofsummary points 272
10.4.1 Fitting the bivariate model in SAS to compare summary points 274
10.4.2 Fitting thebivariate model inStata tocompare summary points 280
10.4.3 Fitting thebivariate model inR tocompare summary points 284
10.4.4 Bayesian inference forcomparing summary points 287
10.4.4.1 Summary statistics 289
10.5 Comparison ofsummary curves 291
10.5.1 Fitting the HSROC model in SAS to compare summary curves 292
10.5.2 Bayesian estimation of the HSROC model for comparing summary curves 294
10.5.2.1 Monitoring convergence 295
10.5.2.2 Summary statistics 295
10.6 Meta- analysis ofsparse data and a typical data sets 296
10.6.1 Facilitating convergence 297
10.6.2 Simplifying hierarchical models 301
10.7 Meta- analysis withmultiple thresholds per study 305
10.7.1 Meta- analysis ofmultiple thresholds withR 306
10.7.2 Meta- analysis ofmultiple thresholds withrjags 311
10.8 Meta- analysis withimperfect reference standard: latent class meta- analysis 316
10.8.1 Specification ofthe latent class bivariate meta- analysis model
inrjags 316
10.8.2 Monitoring convergence 317
10.8.3 Summary statistics and summary ROC plot 317
10.8.4 Sensitivity analyses 320
10.9 Concluding remarks 321
10.10 Chapter information 321
10.11 References 322
xii
11 Presenting findings 327
https://t.me/medicina_free
11.1 Introduction 327
11.2 Results ofthe search 328
11.3 Description ofincluded studies 328
11.4 Methodological quality ofincluded studies 329
11.5 Individual andsummary estimates oftest accuracy 329
11.5.1 Presenting results fromincluded studies 330
11.5.2 Presenting summary estimates ofsensitivity andspecificity 330
11.5.3 Presenting SROC curves 330
11.5.4 Describing uncertainty insummary statistics 332
11.5.5 Describing heterogeneity insummary statistics 333
Comparisons oftest accuracy 333
11.6
11.6.1 Comparing tests using summary points 333
11.6.2 Comparing tests using SROC curves 334
11.6.3 Interpretation ofconfidence intervals fordifferences intest accuracy 336
11.7 Investigations ofsources ofheterogeneity 336
11.8 Re- expressing summary estimates numerically 340
11.8.1 Frequencies 340
11.8.2 Predictive values 341
11.8.3 Likelihood ratios 344
11.9 Presenting findings when meta- analysis cannot beperformed 344
11.10 Chapter information 346
11.11 References 347
Contents
12 Drawing conclusions 349
12.1 Introduction 349
12.2 ‘Summary offindings’ tables 350
12.3 Assessing thestrength ofthe evidence 352
12.3.1 Key issues toconsider when assessing thestrength ofthe evidence 352
12.3.1.1 How valid are thesummary estimates? 359
12.3.1.2
How applicable are thesummary estimates? 359
12.3.1.3 How heterogeneous are theindividual study
estimates? 359
12.3.1.4 How precise are thesummary estimates? 360
12.3.1.5 How complete is thebody ofevidence? 361
12.3.1.6 Were index test comparisons made between or within
primary studies? 362
12.4 GRADE approach for assessing the certainty of evidence 362
12.4.1 GRADE domains for assessing certainty of evidence for test accuracy 363
12.4.1.1 Risk ofbias 363
12.4.1.2 Indirectness (applicability) 363
12.4.1.3 Inconsistency (heterogeneity) 364
12.4.1.4 Imprecision 365
12.4.1.5 Publication bias 365
xiii
Contents
https://t.me/medicina_free
12.5 Summary ofmain results inthe Discussion section 365
12.6 Strengths andweaknesses ofthe review 366
12.6.1 Strengths andweaknesses ofincluded studies 366
12.6.2 Strengths andweaknesses ofthe review 367
12.6.2.1 Strengths andweaknesses dueto thesearch andselection process 367
12.6.2.2 Strengths andweaknesses dueto methodological quality assessment anddata extraction 367
12.6.2.3 Weaknesses dueto thereview analyses 368
12.6.2.4 Direct andindirect comparisons 368
12.6.3 Comparisons withprevious research 369
Applicability offindings tothe review question 369
12.7
12.8 Drawing conclusions 369
12.8.1 Implications forpractice 370
12.8.2 Implications forresearch 373
12.9 Chapter information 374
12.10 References 374
13 Writing aplain language summary 377
13.1 Introduction 377
13.2 Audience andwriting style 378
13.3 Contents andstructure ofa plain language summary 379
13.3.1 Title 380
13.3.2 Key messages 380
13.3.3 ‘Why is improving [...] diagnosis important?’ 381
13.3.4 ‘What is the[...] test?’ 382
13.3.5 What did wewant tofind out? 382
13.3.6 What did wedo? 383
13.3.7 What did wefind? 383
13.3.7.1 Describing theincluded studies 383
13.3.7.2 Presenting information ontest accuracy 384
13.3.7.3 Presenting single estimates ofaccuracy 385
13.3.7.4 Presenting multiple estimates ofaccuracy: two index tests 386
13.3.7.5 Presenting multiple estimates ofaccuracy: more than two index tests 387
13.3.7.6 When presenting anumerical summary oftest accuracy is not appropriate 387
13.3.7.7 Graphical illustration oftest accuracy results 388
13.3.8 What are thelimitations ofthe evidence? 391
13.3.9 How upto date is this evidence? 392
13.4 Chapter information 392
13.5 References 393
13.6 Appendix: Additional example plain language summary 394
Index 399
xiv
Contributors
https://t.me/medicina_free
Arevalo- Rodriguez, Ingrid
Hospital Universitario Ramón y Cajal (IRYCIS) CIBER Epidemiology and Public Health (CIBERESP) Madrid Spain
Bossuyt, Patrick M
Department of Epidemiology and Data Science Amsterdam UMC University of Amsterdam Amsterdam The Netherlands
Chandler, Jacqueline
Wessex Academic Health Science Network Southampton UK
Cumpston, Miranda S
School of Public Health and Preventive Medicine Monash University Melbourne; Cochrane Public Health School of Medicine and Public Health University of Newcastle Newcastle Australia
Davenport, Clare
Institute of Applied Health Research University of Birmingham Birmingham UK
Deeks, Jonathan J
Institute of Applied Health Research University of Birmingham Birmingham UK
Dendukuri, Nandini
McGill University Montreal Canada
Dinnes, Jacqueline
Institute of Applied Health Research University of Birmingham Birmingham UK
Eisinga, Anne
Cochrane UK Oxford University Hospitals NHS Foundation Trust Oxford UK
Flemyng, Ella
Cochrane London UK
Gatsonis, Constantine
School of Public Health Brown University Providence, RI USA
Glanville, Julie
glanville.info York UK
xv
Contributors
https://t.me/medicina_free
Jones, Hayley E
Population Health Sciences Bristol Medical School University of Bristol Bristol UK
Leeflang, Mariska M
Department of Epidemiology and Data Science Amsterdam UMC University of Amsterdam Amsterdam The Netherlands
Li, Tianjing
Department of Ophthalmology School of Medicine University of Colorado Anschutz Medical Campus Aurora, CO USA
Macaskill, Petra
Sydney School of Public Health Faculty of Medicine and Health University of Sydney Sydney Australia
Rücker, Gerta
Institute of Medical Biometry and Statistics Faculty of Medicine and Medical Center University of Freiburg Freiburg Germany
Rutjes, Anne W
Department of Medical and Surgical Sciences SMECHIMAI University of Modena and Reggio Emilia Modena Italy
Schiller, Ian
Centre for Outcomes Research McGill University Health Centre – Research Institute Montreal Canada
Scholten, Rob J
Cochrane Netherlands Julius Center for Health Sciences and Primary Care University Medical Center Utrecht Utrecht University Utrecht The Netherlands
Partlett, Christopher
Nottingham Clinical Trials Unit University of Nottingham Nottingham UK
Reitsma, Johannes B
Julius Center for Health Sciences and Primary Care University Medical Center Utrecht Utrecht University Utrecht The Netherlands
xvi
Spijker, René
Cochrane Netherlands Julius Center for Health Sciences and Primary Care University Medical Center Utrecht Utrecht University Utrecht; Medical Library Amsterdam UMC University of Amsterdam Amsterdam The Netherlands
Contributors
https://t.me/medicina_free
Steingart, Karen R
Department of Clinical Sciences Liverpool School of Tropical Medicine Liverpool UK
Takwoingi, Yemisi
Institute of Applied Health Research University of Birmingham Birmingham UK
Whiting, Penny
Population Health Sciences Bristol Medical School University of Bristol Bristol UK
Yang, Bada
Julius Center for Health Sciences and Primary Care University Medical Center Utrecht Utrecht University Utrecht The Netherlands
xvii
Preface
https://t.me/medicina_free
Patrick M. Bossuyt, Jonathan J. Deeks, Mariska M. Leeflang, Yemisi Takwoingi and Ella Flemyng
Medical tests are an indispensable element of modern- day health care. Clinicians rely on medical tests to find the likely cause of a patient’s signs and symptoms, to evaluate the extent of disease, to predict the future course of the condition, to screen for asymp­tomatic disease and in many other situations.
Like any other intervention in health care, a medical test should be properly evalu­ated before its use can be recommended. Tests must make little error in measuring chemical, biological or physical quantities, or in detecting features, whether a clinical doctor assessing symptoms, a pathologist seeing a biopsy or a radiologist identifying an image. However, as well as being accurate, tests must also provide the right informa­tion to guide clinical actions. Forthis reason, a medical test must be evaluated for its clinical performance.
For diagnostic tests, this clinical performance is referred to as diagnostic accuracy: the ability of a test to correctly identify people who have or do not have the target con­dition. A diagnostic accuracy study assesses the results of an index test (test of interest) against a reference standard and provides estimates of the test’s performance. Such evaluations of clinical performance should be done in a real- world setting within the context of a clinical pathway taking into account the intended use and target population.
For many tests a range of diagnostic accuracy studies have been reported in the medical literature. As in other areas of science, systematic reviews of such studies can provide informative syntheses of the available evidence. A systematic review can provide decision makers and other stakeholders with an overview of the currently avail­able evidence about a test’s diagnostic accuracy. We expect such reviews to include a well- defined review question, systematic searches of all relevant studies, evaluations of
This chapter should be cited as: Bossuyt PM, Deeks JJ, Leeflang, MM, Takwoingi Y, Flemyng E. Preface. In: Deeks JJ, Bossuyt PM, Leeflang MM, Takwoingi Y, editors. Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy. 1st edition. Chichester (UK): John Wiley & Sons, 2023: xix–xxiv.
xix
Preface
https://t.me/medicina_free
the risk of bias and applicability of identified studies and, if possible, meta- analyses that statistically combine the results of multiple studies.
The methods for systematic reviews of diagnostic accuracy studies were slow to develop, compared to methods for systematic reviews of randomized controlled trials of interventions. This difference parallels a difference in rigour in the primary studies and an understanding of sources of bias. Other explanations can be found as well: the wide differences in study objectives and designs of test accuracy studies, ranging from bio­marker discovery studies to large- scale clinical applications in the intended use setting. This variation and complexity of study designs can lead to results that are either not rel­evant or not valid for answering a test accuracy review question. This is further com­pounded by the fact that measures of test accuracy are not automatically transferable across different populations and settings. Unlike randomized controlled trials, test accu­racy studies typically provide paired proportions to indicate how well the test performs in those who have and those who do not have the target condition, i.e. the test’s sensitiv­ity and specificity. This poses specific challenges for meta- analysis. Furthermore, sensi­tivity and specificity vary considerably more across studies than estimates of relative risk from clinical trials because they are proportions, not relative or absolute differences.
“Another challenge is in addressing clinically important comparative questions when there are competing tests that can be used at the same point in the clinical path­way. Systematic reviews of test accuracy can evaluate and compare the accuracy of two or more tests. While many studies evaluate the accuracy of a single index test, far fewer compare the accuracy of two or more index tests. Such comparative test accuracy stud­ies, which can provide robust evidence to inform test selection, deserve greater appre­ciation from clinical investigators, researchers, grant- awarding organizations funding test research, and those developing test use recommendations.”
“Evidence on how the results of primary studies differ according to methodological features may be provided by meta-epidemiological studies. These meta-epidemiologi­cal studies often use data collected from systematic reviews. For example, the earliest meta-epidemiological studies based on Cochrane reviews strengthened the impor­tance of allocation concealment in randomised trials. Similar empirical studies using systematic reviews of diagnostic accuracy have confirmed essential elements of valid test accuracy studies, from avoiding the unnecessary inclusion of healthy controls to the need for appropriate statistical methods.”
“This Handbook is informed by currently available empirical evidence and theoretical understanding of primary diagnostic accurarcy studies, review methods, and tech­niques for meta-analysis. The ongoing creation of Cochrane reviews will provide data for future studies that will shape our knowledge of how primary test accuracy studies and systematic reviews could be better performed.”
This Handbook has a long history. Its development started around 2002, when a num­ber of us felt that the methodology for systematic reviews of test accuracy studies had sufficiently advanced to guide researchers. Since then, an incredible number of col­leagues have contributed to its gestation. Many of these have become authors of chap­ters in this Handbook. Others are explicitly acknowledged in various chapters for their contribution. Yet many more have contributed, through various discussions in methods groups sessions, conferences and other meetings. While we worked on this Handbook progress did not halt and in several steps of the review process more methodological advances were made, while best practices emerged in other areas.
xx
Preface
https://t.me/medicina_free
Considerable progress has been made in searching, in evaluating risk of bias and appli­cability and in meta- analysis. The continuing development of software programs and user­written macros has made meta- analysis methods more accessible. A challenging area remains the communication of findings from reviews to decision makers and other stake­holders. Systematic reviews typically reveal substantial variability across studies and this variability makes it challenging to appreciate the true diagnostic accuracy of medical tests.
The authors of this Handbook have aimed to develop guidance based on the best current knowledge. We expect that several of the methods and practices will be advanced even further in the coming years, necessitating an update of this Handbook in the future. We are confident that the guidance in this Handbook will enable researchers in Cochrane and beyond to prepare informative reviews of the available literature to benefit patients and clinicians. Inshort: to all of us.
About Cochrane
Cochrane is an international network of health practitioners, researchers, methodolo­gists, patients and carers, and others, with a vision of a world of better health for all people where decisions about health and care are informed by high- quality evidence (www.cochrane.org). Founded as The Cochrane Collaboration in 1993, it is a not- for­profit organization whose members aim to produce credible, accessible health informa­tion that is free from commercial sponsorship and other conflicts of interest.
Cochrane works collaboratively with health professionals, policy makers and interna­tional organizations, such as the World Health Organization (WHO), to support the development of evidence- informed guidelines and policy. WHO guidelines on critical public health issues such as the consolidated guidelines on systematic screening for tuberculosis disease (2021) and rapid diagnostics for tuberculosis detection (2021), and the WHO Essential Diagnostics List (2021), are underpinned by Cochrane Reviews.
There are examples of the impact of Cochrane Reviews on health and health care. Globally in 2020, a WHO- recommended rapid molecular test was used as the initial diag­nostic test for 1.9million (33%) of the 5.8million people newly diagnosed with tubercu­losis in 2020 (Global Tuberculosis Report 2021). Several rapid diagnostic tests (RDTs) for tuberculosis have been endorsed by the WHO since 2010 based on evidence from many Cochrane Reviews. A Cochrane Special Collection on diagnosing tuberculosis, first published in 2019 and last updated in March 2022 to celebrate World Tuberculosis Day (www.cochranelibrary.com/collections/doi/SC000034/full), highlighted the influential Cochrane reviews that have informed the two WHO consolidated guidelines on system­atic screening and rapid diagnostics for tuberculosis. The use of RDTs for tuberculosis has contributed to the decentralization of testing and allowed patients to be diagnosed quickly with earlier initiation of appropriate treatment in many low- and middle- income countries. Rapid and reliable testing has also facilitated earlier detection of drug resist­ance and reduced mortality from tuberculosis among persons living with HIV.
Cochrane Reviews are published in full online in the Cochrane Database of Systematic Reviews, which is a core component of the Cochrane Library (www.cochranelibrary. com). The Cochrane Library was first published in 1996, and is now an online collection of multiple databases. The first Cochrane Review of diagnostic test accuracy was pub­lished in 2008.
xxi