The State Food and Drug Administration issued technical guidelines for clinical trials of new in vitro diagnostic reagents

On September 27, in order to guide the clinical trials of in vitro diagnostic reagents, the State Food and Drug Administration issued technical guidelines for in vitro diagnostic reagents clinical trials, which will be implemented from the date of issuance.
"Technical Guiding Principles" shall be repealed
.
The original text is as follows: Announcement of the State Food and Drug Administration on Issuing Technical Guidelines for In Vitro Diagnostic Reagents (No.
72 of 2021) In order to guide clinical trials of in vitro diagnostic reagents, in accordance with the "Administrative Measures for the Registration and Filing of In Vitro Diagnostic Reagents" (National Market Supervision) The State Administration of Management Order No.
48), the State Food and Drug Administration has organized and formulated the "Guiding Principles for Clinical Trials of In Vitro Diagnostic Reagents" (see appendix), which is hereby issued
.
The technical guidelines will be implemented from the date of issuance, and the "Technical Guidelines for Clinical Trials of In Vitro Diagnostic Reagents" issued by the former State Food and Drug Administration (formerly CFDA Notice 2014 No.
16) shall be repealed
.
Annex Technical guidelines for clinical trials of in vitro diagnostic reagents 1.
Scope of application Clinical trials of in vitro diagnostic reagents refer to a systematic study of the clinical performance of in vitro diagnostic reagents in the corresponding clinical environment
.
The purpose of clinical trials is to prove that in vitro diagnostic reagents can meet the requirements of the intended use, and to determine the applicable population and indications of the product
.
The results of clinical trials provide effective scientific evidence for the confirmation of the safety and effectiveness of in-vitro diagnostic reagents and risk-benefit analysis
.
In vitro diagnostic reagents provide test results through in vitro detection of human samples, which are used alone or together with other information to assist in judging the subject's target state (health state, disease state, disease process, or other disease/health state that can be used to guide clinical treatment, etc.
) )
.
The "clinical performance" of the in vitro diagnostic reagent refers to the ability of the in vitro diagnostic reagent to be used by the intended user in the expected use environment, and to obtain test results related to the target state of the subject for the target population
.
This guideline is applicable to clinical trials conducted in China in accordance with in vitro diagnostic reagents managed by medical devices and used for registration applications in China
.
This guideline aims to clarify the basic principles of clinical trials and the key factors that need to be considered in clinical trials, and put forward basic requirements for clinical trial quality management, which are used to guide the clinical trial work of sponsors and also provide technical review departments for clinical trials.
The review of the data provides reference
.
Since in vitro diagnostic reagent products have the characteristics of rapid development, large professional span, and different clinical expected uses, the clinical test methods and contents of different products are not the same
.
The sponsor should formulate a reasonable clinical trial plan based on the specific conditions of the product.
The content of this guideline will also be revised in due course according to the needs of the development of in vitro diagnostic reagents
.
2.
Basic principles (1) Ethical principles Clinical trials should follow the ethical guidelines of the "World Medical Congress Declaration of Helsinki" and relevant national requirements for biomedical research ethics involving humans, and should be reviewed and approved by the ethics committee
.
Researchers need to consider the acquisition of samples for clinical trials, such as blood, urine, sputum, cerebrospinal fluid, feces, vaginal secretions, nasopharyngeal swabs, tissue sections, bone marrow, amniotic fluid, etc.
, and the risks to subjects of the test results.
Request the ethics committee to review to ensure that clinical trials will not put subjects at unreasonable risks, and obtain informed consent from subjects (or their guardians) as required
.
(2) Scientific principles The development of clinical trials should be established on the basis of pre-clinical research, with sufficient scientific basis and clear trial objectives
.
According to the expected use of the product, the epidemiological background of related diseases, and statistical requirements, the clinical trials should be scientifically designed, while the test errors should be controlled to the greatest extent, the test quality should be improved, and the test results should be scientifically and reasonably analyzed
.
While ensuring that the test results are scientific, accurate, and credible, they should be as efficient, fast and economical as possible
.
(3) The principle of law.
This guiding principle is set out in the "Regulations on the Supervision and Administration of Medical Devices" (Order No.
739 of the State Council), the "Administrative Measures for the Registration and Filing of In Vitro Diagnostic Reagents" (Order No.
48 of the State Administration for Market Regulation) and the "Medical Devices Developed under the regulatory framework of "Clinical Trial Quality Management Standards"
.
The development of clinical trials of in vitro diagnostic reagents shall comply with the requirements of relevant laws and regulations
.
1.
Clinical trial institutions and personnel Clinical trials of in vitro diagnostic reagents should be carried out in clinical trial institutions for medical devices that have the appropriate conditions and are filed in accordance with regulations
.
Clinical trials of in vitro diagnostic reagents should be carried out in multiple clinical trial institutions in accordance with the same clinical trial protocol
.
The sponsor is responsible for selecting and determining the coordinating investigator for clinical trials of in vitro diagnostic reagents, and the medical institution where the coordinating investigator works is the team leader unit
.
Coordinating researchers undertake the coordination work of various centers in clinical trials
.
The ethics committee of the leader unit is responsible for reviewing the ethics and scientificity of the trial protocol.
The ethics committees of other clinical trial institutions participating in the trial will review the feasibility of the trial in this clinical trial institution under the premise of accepting the review opinions of the ethics committee of the lead unit.
Including the qualifications and experience of the investigator, equipment and conditions, etc.
, under normal circumstances, no amendments to the design of the trial plan will be proposed, but the trial has the right not to approve the trial in its clinical trial institution
.
In-vitro diagnostic reagent clinical trial institutions shall have the professional technical level, organizational and management capabilities required for clinical trials, and be able to carry out ethical review work, and have the personnel, facilities and conditions suitable for the clinical trials to be carried out
.
This includes but is not limited to: routinely carry out relevant test items and/or disease diagnosis and treatment items, have the ability to interpret relevant diagnosis results and disease treatment, and have the emergency mechanism and handling capabilities to prevent and deal with emergencies and serious adverse events in clinical trials; has to meet the needs of clinical trial study population; laboratory with the necessary conditions to meet the relevant testing laboratory accreditation requirements (if any) and the like
.
The clinical trial institution should be able to ensure that the clinical trial is implemented in strict accordance with the plan, and can cooperate with the product registration application process, including conducting necessary supplementary trials, cooperating with the supervision and inspection organized by the sponsor, and conducting inspections by the drug supervision and management department and the health management department.
Check etc. .
The principal investigator of clinical trials should have the ability to design and implement relevant clinical trials, have the professional knowledge and experience required for clinical trials of in vitro diagnostic reagents, and should be familiar with the requirements of relevant clinical trial regulations
.
After training, the personnel participating in the clinical trial should be familiar with the principles, scope of application, operation methods, etc.
of the relevant detection technology, and be able to correctly interpret the test results
.
The person in charge of clinical trial statistics should be a person with relevant professional background and professional ability
.
2.
Clinical trial plan and report 2.
1 Clinical trial plan To carry out clinical trials of in vitro diagnostic reagents, the sponsor should organize the development of scientific and reasonable clinical trials according to the purpose of the trial, comprehensively considering the expected use, product characteristics and expected risks of the test in vitro diagnostic reagents Scheme
.
The clinical trial protocol should be strictly followed during the entire clinical trial after being approved by the ethics committee
.
Each clinical trial institution shall implement the same clinical trial plan.
The plan shall make clear provisions on the type of trial design, selection of comparison methods, selection of subjects, evaluation indicators, statistical analysis methods, sample size estimation and quality control requirements, etc.
, and shall be based on The situation of each institution reasonably determines the sample size allocation plan
.
For specific requirements on the content of the clinical trial program, please refer to the "Medical Device Clinical Trial Quality Management Specification" and its annex in vitro diagnostic reagent clinical trial program template
.
2.
2 Clinical trial summary and report After the clinical trial, the clinical trial data of each clinical trial institution should be summarized separately, and the clinical trial summary should be issued, together with the clinical trial data sheet, other test methods used in the clinical trial or other in vitro diagnostic reagents and other products Basic information and other materials
.
The text of the clinical trial summary should include the following: 2.
2.
1 Overview of the clinical trial (in vitro diagnostic reagent information, test flow chart, comparison method information, other relevant testing methods/reagent information, subject enrollment, sample collection and storage, Testing equipment, sample size included in statistics, description of sample removal, summary and descriptive analysis of test data, etc.
); 2.
2.
2 Version number and version date of the implementation plan, clinical trial personnel information, trial start time and end time; 2.
2 .
3 Clinical trial quality management, including: pre-clinical trial training, clinical trial process quality control, bias control measures, record management, data management, sample, reagent and instrument management, etc.
; 2.
2.
4 Description of clinical trial ethics ; 2.
2.
5 Occurrence and handling of adverse events and device defects; 2.
2.
6 Description of plan deviation and plan modification; 2.
2.
7 Others
.
Clinical trial data table includes at least: The only traceable sample number, demographic information (gender, age), the clinical diagnosis of subjects background information, sample type, test results and so on
.
Clinical trials required when attached original map and the like
.
The clinical trial data sheet should be signed by the clinical trial operator and reviewer, and the clinical trial institution's signature (cover seal and seam seal)
.
The sponsor and coordinating investigator will complete the clinical trial report after summarizing and statistically analyzing all clinical trial data
.
For specific requirements on the content of the clinical trial report, please refer to the "Medical Device Clinical Trial Quality Management Practice" and its annex in vitro diagnostic reagent clinical trial report template
.
2.
3 The clinical trial plan and clinical trial summary should be signed and dated by the principal investigator, and submitted to the sponsor after being reviewed and signed by the medical device clinical trial institution
.
The clinical trial report shall be signed and dated by the coordinating investigator, and submitted to the sponsor after being reviewed and signed by the medical device clinical trial institution of the team leader
.
All data of the clinical trial should be signed and sealed by the sponsor
.
3.
Clinical trial design In vitro diagnostic reagent clinical trial design and product intended use, indications, applicable population (target population), characteristics of the test substance, test sample type, product use method (such as user) and test result reporting method (such as Qualitative, quantitative) and so on are directly related
.
The conclusion of the clinical trial should be able to prove that the clinical performance of the product meets the requirements of the intended use and support the relevant content described in the instructions
.
According to product characteristics and product performance evaluation needs, clinical trials of in vitro diagnostic reagents may include different clinical trial purposes.
It is necessary to conduct scientific clinical trial design for each clinical trial purpose, including selecting appropriate clinical trial design types and determining the appropriate ones.
Comparison methods, inclusion/exclusion criteria of subjects, and clinical evaluation indicators, etc.
, and scientific sample size estimation
.
At the same time, clinical trials are a process of obtaining research results based on a limited sample of subjects obtained by sampling, and making statistical inferences about the population of subjects (target population) with similar conditions in the future
.
Therefore, in the design of clinical trials, it is necessary to make reasonable and effective arrangements for the relevant factors of the trial based on statistical principles, and conduct a scientific and reasonable analysis of the trial results
.
For brand-new in vitro diagnostic reagents or products that are significantly different from similar products on the market, before the formal clinical trial is launched, a small sample exploratory test can be considered
.
The development of exploratory trials helps to reduce the possibility that unexpected results will lead to the need to change key designs in clinical trials
.
Generally speaking, in order to make scientifically valid corroborating reasoning, exploratory trial data should not be combined with research data in the clinical trial phase
.
This article does not put forward specific requirements for the design and management of exploratory experiments, but the relevant content of this article can be used for reference in the process of exploratory experiment implementation
.
(1) Types of design According to the impact of the test results of in vitro diagnostic reagents on subjects during clinical trials, clinical trials of in vitro diagnostic reagents mainly include two design types: observational studies and interventional studies
.
In observational studies, while testing the samples with test in vitro diagnostic reagents, subjects will also receive routine clinical diagnosis and laboratory testing.
The test results of in vitro diagnostic reagents are not used for patient management and do not affect clinical decision-making; clinical trials Confirm the clinical performance of the product by evaluating the consistency of the test result with the clinical reference standard (or other method) for determining the target state of the subject
.
In interventional studies, the test results of in-vitro diagnostic reagents will be used for patient management or to guide treatment, and by evaluating treatment effects or patient benefits, it provides evidence to support the determination of the safety and effectiveness of in-vitro diagnostic reagents
.
When designing clinical trials, the appropriate design type should be selected according to the characteristics and intended use of in vitro diagnostic reagents
.
The content of this guideline is mainly based on the discussion and requirements of observational research design, and other clinical trial design situations can be implemented based on specific conditions
.
1.
Cross-sectional studies and longitudinal studies in observational research are generally, the observational studies of in vitro diagnostic reagents mainly involve cross-sectional studies, that is, to evaluate the test results of samples collected at a single time point and the judgment results of clinical reference standards (or other methods) Consistency
.
In addition, some products require longitudinal data research, that is, the test results of samples collected at multiple time points can be used to evaluate the clinical performance of the product
.
For example, for certain in vitro diagnostic reagents used for treatment monitoring, in clinical trials, subjects and the test substances in their samples should be observed at multiple time points before and after treatment to prove the changes in test results and the condition of the test substances.
The correlation between development and treatment effect
.
The clinical trial plan should clarify the observation time of the subject and the clinical evaluation index according to the characteristics of the test object and the disease process
.
2.
Selection of comparison methods in observational studies.
Under normal circumstances, in observational studies, in vitro diagnostic reagents and clinical reference standards are used for comparative research to evaluate the correlation between the test results of in vitro diagnostic reagents and the target state of subjects.
Evaluation indicators generally include clinical sensitivity and clinical specificity, etc. .
Clinical reference standards refer to the best clinically available methods that can be used to determine the target state of subjects under existing conditions, usually from clinical and laboratory medical practice, including: recognized, reliable, and reliable under existing conditions Authoritative disease diagnosis standards (such as histopathological examinations, imaging examinations, pathogen isolation and culture identification, conclusions obtained from long-term follow-up, etc.
), disease diagnosis methods specified in the disease diagnosis and treatment guidelines, recommended by the consensus of experts in the industry or clinically recognized reasonable reference methods
.
The clinical reference standard may be one method or a combination of multiple methods
.
For in-vitro diagnostic reagents that have similar products on the market in China, clinical trials can also use the method of comparing test in-vitro diagnostic reagents with similar products on the market (comparative reagents) to evaluate the consistency of the test results of the two methods.
The evaluation indicators usually include The positive coincidence rate, negative coincidence rate
.
In contrast agents intended use, for the crowd aspect, the type of sample, detection methodologies, detection performance, etc.
should have a good comparability with the test in vitro diagnostic reagents
.
In order to evaluate the clinical performance of in vitro diagnostic reagents more comprehensively, it is sometimes necessary to combine comparative studies with clinical reference standards and comparative studies with similar products on the domestic market in clinical trials to comprehensively evaluate the clinical performance of the products to support relevant All claims for intended use
.
In the case that there is no clinical reference standard, or the clinical reference standard cannot fully evaluate the clinical performance of the product, and there is no similar product on the market in China, when designing the clinical trial plan, the clinical significance of the product should be proved At the same time, based on the existing clinical practice and theoretical basis, we choose currently recognized and reasonable methods to conduct comparative studies to further confirm the clinical performance of the product
.
The selection of clinical trial comparison methods should be comprehensively considered based on the expected use of the product, the type of sample, the method of reporting test results, and the availability of clinical reference standards and comparison reagents.
The conclusion of the clinical trial should be able to support the content of the intended use claim
.
The basis for the selection of the comparison method should be described in the clinical trial protocol
.
Clinical trials of in vitro diagnostic reagent change registration can use the products after the change and the products before the change for comparative research; for the situation where the product performance changes significantly before and after the change or the clinical indications are added, the products after the change and the clinical reference standards or the domestic existing products can also be used.
A comparative study of similar products on the market proved that the clinical performance of the changed product meets the requirements
.
3.
Special situations in observational research.
For some in vitro diagnostic reagents, there may be situations that require special consideration in the design of clinical trials.
If the difference is such that the same sample cannot be used for testing (for example: the applicable sample is a swab sample, but the swab material and preservation solution applicable to the two methods are different), at this time, two samples can be taken for each subject Collect and test the in vitro diagnostic reagents and contrast reagents separately.
The order of the two samples should follow the random principle
.
It should be noted that this test method is generally only considered when one sample collection will not affect the next sample collection
.
(2) Subject selection and sample collection The clinical trial plan should reasonably determine the inclusion/exclusion criteria of clinical trial subjects, and the stratification of subjects into groups based on the intended use of the test in vitro diagnostic reagents, target population, and testing requirements, etc.
requirements and sample collection methods
.
1.
Subjects of clinical trials Clinical trial subjects should come from the applicable population (target population) claimed by the intended use of the product, such as those with certain symptoms, signs, physiology, pathological conditions, or certain epidemiological backgrounds, etc.
People
.
Enrollment of non-target populations may introduce subject selection bias, resulting in clinical trial results that cannot reflect the true situation of the product
.
The test population should be able to represent the characteristics of the target population, including demographic characteristics (gender, age), symptoms, signs, comorbidities, and the stage, location and severity of the disease; at the same time, the subjects should exclude those who are not suitable for the clinical trial Physiological or pathological characteristics
.
According to the above requirements, the inclusion/exclusion criteria of subjects shall be set reasonably, and appropriate measures shall be taken during the clinical trial process to ensure that only those who meet the criteria can be included in the group
.
In addition, subjects need to consider other possible influencing factors according to product characteristics, such as the influence of different ethnic groups, different races, and different regions
.
For example, for products used for disease-assisted diagnosis and differential diagnosis, the subjects should come from people with suspected symptoms or related epidemiological backgrounds, including subjects with the target disease state and subjects without the target disease state.
Examiner
.
Subjects with the target disease state should be able to cover all the characteristics of the disease state as much as possible, including typical and atypical symptoms, disease type, stage, length of disease, severity of disease, etc.
, in order to evaluate the clinical sensitivity of the product; Subjects with the target disease state need to include other disease cases with the same or similar symptoms, easily confused with the target disease state, etc.
, to evaluate the clinical specificity of the product; in addition, consideration should be given to including samples that may interfere with the detection, etc.

.
2.
Subjects are enrolled by stratification.
When the clinical performance of in vitro diagnostic reagents is expected to be different in different subgroups, and the clinical performance of some important subgroups needs to be accurately evaluated, it is recommended to adopt stratified enrollment.
, And the sample size of the subgroup should meet statistical requirements
.
Stratified enrollment is to divide the target population into pre-designated non-overlapping different subgroups, and enroll subjects separately for each subgroup
.
For example, stratify the target population by gender (male, female) and age group (below or higher than a certain age) as needed
.
The stratified entry method not only ensures the full evaluation of important subgroups, but also helps to obtain more accurate performance results
.
3.
Sample collection In the clinical trials of in vitro diagnostic reagents, subjects should be enrolled and samples should be collected in accordance with the enrollment/exclusion criteria, subject recruitment methods, and sample collection methods specified in the clinical trial protocol
.
Under certain circumstances, some subjects’ samples can come from past, other research sample sets, or sample sets for no specific purpose
.
In this case, special attention should be paid to avoid introducing bias.
For example: ①During the test process, after blinding, the test operator and the result reader should not be able to distinguish the samples from the previous sample set; ②The storage and handling of the previous sample set Etc.
should meet the requirements; ③ The inclusion of samples from previous sample sets may cause the proportion of subjects with rare conditions in the clinical trial population to be significantly higher than the proportion of the target population in the natural state.
This potential bias should be calculated Take it into consideration in the analysis
.
If a large number of previous samples are included, special attention should be paid to fully demonstrate possible selection bias and other issues, such as: ① Whether the test population can represent various characteristics of the target population (not just the most typical characteristics); ② Whether the sample comes from enough A large sample set can meet the randomness requirements of sampling to a certain extent; whether the sample in the sample set has non-random sample removal; ③whether the sample has sufficient clinical information of the subject; ④whether the sample is in the clinical trial of quantitative detection It can cover the entire linear range; ⑤Whether the sample is a suitable sample type for the product, and the storage conditions and time meet the stability requirements of the measured object
.
(3) Number and requirements of clinical trial institutions 1.
Number of clinical trial institutions For in vitro diagnostic reagents that require clinical trials, no less than 2 (including 2) clinical trial institutions that meet the requirements should be selected for the second category of products.
For similar or newly developed products, no less than 3 (including 3) clinical trial institutions that meet the requirements should be selected to carry out clinical trials
.
If it is necessary to change the registration of clinical trials, generally no less than 2 (including 2) clinical trial institutions that meet the requirements can be selected to carry out clinical trials
.
2.
Clinical trial institutions require sponsors to select representative institutions to carry out clinical trials based on the characteristics of the product and its intended use, as well as factors such as population differences in different regions, epidemiological background, and characteristics of pathogenic microorganisms, including representatives of the tested population Characteristics, representativeness of clinical conditions (expected use environment and users), etc. .
3.
Central effect.
The entire clinical trial process should be carried out in strict accordance with the clinical trial protocol and relevant standard operating procedures approved by the ethics committee.
The trial operation and result interpretation should be consistent with the relevant regulations in the in vitro diagnostic reagent product manual; the clinical trial started Before, the sponsor should be responsible for organizing training related to the clinical trial, including the storage, operation, and management of test in vitro diagnostic reagents, to ensure the consistency of clinical trial operations and maximize control of test errors; Necessary monitoring and quality control measures; the sample size of each clinical trial institution should be balanced as much as possible; the possible impact of the different clinical trials carried out by each clinical institution should be considered, and clinical trials should be carried out at the same time as possible
.
It should be noted that even though the same research protocol is used, when the clinical data of various clinical trial institutions are combined, there may still be a central effect, that is, significant differences in the results of the trials of each center
.
The central effect may reflect the differences in the test population and clinical trial operation among the centers, and the generation of the central effect may affect the interpretation of the research results
.
When designing clinical trial protocols, consideration should be given to how to avoid potential bias due to central effects
.
(4) Clinical evaluation index The clinical evaluation index should be determined during the design phase of the clinical trial and be clarified in the clinical trial plan
.
In vitro diagnostic reagent clinical test evaluation indicators for qualitative testing usually include diagnostic accuracy (sensitivity, specificity, predictive value, likelihood ratio, area under the ROC curve, etc.
) or test consistency (positive/negative coincidence rate, total coincidence rate, Kappa Value, etc.
)
.
Evaluation of clinical trials semi-quantitative determination in vitro diagnostic reagents typically include: various grades coincidence rate, female / positive coincidence rate and Kappa values and the like
.
Evaluation indicators for clinical trials of in vitro diagnostic reagents for quantitative detection usually include regression coefficients, intercepts, correlation coefficients, and determination coefficients of regression analysis. .
(5) Statistical analysis of clinical trials The statistical analysis of clinical trial results should be based on correct and complete data.
Appropriate clinical evaluation indicators should be selected to evaluate the clinical performance of in vitro diagnostic reagents, and appropriate statistical models should be used to analyze the data.
Analysis
.
1.
Basic considerations of statistical analysis Statistical analysis of in vitro diagnostic reagents generally includes parameter estimation (including confidence interval estimation) and hypothesis testing of evaluation indicators
.
Parameter estimation is to confirm the sensitivity, specificity, regression coefficient (of regression equation), intercept and other evaluation indexes under the premise of ensuring that the evaluation indexes meet the expected accuracy level (the width of the confidence interval is certain)
.
Hypothesis testing needs to put forward null hypotheses and alternative hypotheses for statistical indicators, and make relevant statistical inferences through hypothesis testing
.
If necessary, it is necessary to consider verifying the distribution hypothesis of the sample before statistical analysis, so as to choose a reasonable statistical method
.
For studies with definite clinical acceptance criteria, the acceptance criteria should be clearly defined in the clinical trial protocol, and the statistical analysis of clinical trial data proves that the evaluation results (interval estimation) of the clinical evaluation indicators meet the requirements of the acceptance criteria
.
Clinical acceptance standards should be the results widely recognized by the industry, and are generally set based on factors such as the performance level of relevant test reagents, risk determination, and clinical needs
.
1.
1 Statistical analysis of qualitative testing Qualitative testing clinical trials generally summarize the test results of the two analytical methods in the form of 2×2 tables, and calculate the sensitivity (positive coincidence rate), specificity (negative coincidence rate), and total coincidence rate based on this , Kappa value and other indicators and their confidence intervals
.
In addition, hypothesis testing can also be performed at the same time to evaluate the consistency of the two analysis methods
.
1.
2 Statistical analysis of semi-quantitative detection In vitro diagnostic reagents for semi-quantitative detection usually refer to: the test results are reported as several grades (for example: negative, +, 2+, 3+), or reagents that are reported as the end-point dilution, etc. .
In clinical trials, the test results can be summarized in the form of R×C table, and the coincidence rate, negative/positive coincidence rate, Kappa value and other indicators and their confidence intervals of each grade can be calculated based on this
.
1.
3 Statistical analysis of quantitative detection 1.
3.
1 Scatter diagrams should be drawn based on clinical trial data, and correlation analysis should be performed
.
1.
3.
2 The Bland-Altman method is used to calculate the consistency limit and evaluate the consistency of the two test results
.
The consistency limit should be within the range of clinically acceptable limits
.
1.
3.
3 Use regression analysis to evaluate the consistency of the two detection methods
.
Applicable regression analysis methods should be selected according to factors such as data distribution characteristics, such as Passing-Bablok regression, Deming regression, and least square regression
.
Regression analysis should focus on observing the regression coefficient and intercept of the regression equation and calculating the confidence interval of the regression coefficient and intercept
.
It is also possible to conduct hypothesis testing on related evaluation indicators at the same time
.
1.
3.
4 Special analysis should be made for the test results near the medical decision level
.
1.
4 ROC analysis For the test results of in-vitro diagnostic reagents as quantitative or semi-quantitative data, the clinical reference standard judgment result is a statistical analysis of qualitative results.
The receiver operating characteristic (ROC) curve method can also be used to take the area under the ROC curve Reflect the diagnostic value of the test in vitro diagnostic reagents, or compare the diagnostic value of two reagents at the same time
.
For clinical trials of in vitro diagnostic reagents, when ROC analysis is used for data statistics, the recommended positive judgment value should be further evaluated for sensitivity, specificity and other indicators (and their confidence interval)
.
2.
Analysis of inconsistent samples for qualitative testing.
In the clinical trials of qualitative testing reagents, if there is any inconsistency between the test results of in vitro diagnostic reagents and comparison methods, a comprehensive analysis of the inconsistent results should be carried out to indicate whether it affects the judgment of the product's clinical performance; Samples with inconsistent results can be analyzed using clinical reference standards or other appropriate methods, but the analysis results should not be included in the original statistical analysis
.
3.
If necessary, the obtained data set should be stratified and segmented for statistics
.
4.
Samples included in clinical trials should not be arbitrarily eliminated.
Sample elimination standards should be set in the clinical trial protocol; if any, the circumstances of all sample eliminations should be detailed in the clinical trial summary and report, and the reasons should be explained
.
5.
In the whole process of clinical trials of in vitro diagnostic reagents, the conclusions of biostatistical analysis are very important.
At the same time, the requirements of clinical diagnosis and treatment for the clinical performance of in vitro diagnostic reagents should be fully considered; in the evaluation of product safety and effectiveness, comprehensive Consider the statistical significance of the evaluation results and the clinical significance they represent
.
(6) Requirements of sample size Appropriate sample size is a necessary condition to ensure accurate evaluation of the clinical performance of in vitro diagnostic reagents
.
The sample size of clinical trials of in vitro diagnostic reagents is related to a variety of factors, including evaluation indicators, test repeatability, interference factors, differences between subgroups, and characteristics of the test substance
.
In the clinical trial plan, the minimum sample size required for the clinical trial should be estimated and the basis should be stated
.
1.
Basic considerations about sample size requirements 1.
1 The sample size of clinical trials should meet the statistical requirements, and appropriate statistical methods should be used for estimation
.
1.
2 The determination of clinical trial sample size should consider various influencing factors of clinical performance to ensure adequate evaluation of clinical performance, such as: the test population should be able to represent various characteristics of the target population, taking into account the evaluation of detection performance in different subgroups Needs, as well as the requirements for testing performance evaluation of multiple test objects (or multiple subtypes, etc.
), the minimum total sample size should be estimated to ensure that the number of samples of various groups/types meets the requirements
.
When the clinical performance of in vitro diagnostic reagents is expected to be different in different subgroups, and the clinical performance of certain important subgroups needs to be accurately evaluated, the sample size of the subgroups should be separately statistically estimated
.
1.
3 If the test in vitro diagnostic reagent is suitable for different sample types, the sample size requirements of different sample types need to be considered
.
When different sample types are used to detect the same analyte, the analysis performance, clinical performance, and applicable population of the product may be caused by factors such as the difference in the sample matrix, the source of the analyte, the difference in the concentration level of the analyte, the difference in interference factors, and the difference in the sampling location.
, differences in indications, like reference interval
.
For different sample types, the impact of the above-mentioned differences should be considered, and reasonable sample size requirements should be determined
.
If different sample types have significant differences in clinical performance, applicable populations, indications, reference intervals, etc.
, clinical trial design should be carried out separately for different sample types, including sample size estimation and statistical analysis
.
If pre-clinical research proves that there are differences in analytical performance between different sample types, but does not have a significant impact on clinical performance, such as serum and whole blood samples in certain immunological tests, clinical trials should be based on one type of sample For clinical trials, the number of sample cases meets the statistical requirements; other sample types are appropriately included in a certain number of positive and negative samples, and the clinical performance is evaluated through a homologous sample comparison test or a comparison study with a comparison method
.
For example, different sample types have almost no differences in sample matrix, analyte source, analyte concentration level, interference factors, and sampling location, and pre-clinical studies have proved that there is no difference in analytical performance, such as serum and plasma in some tests.
Sample, different sample types in clinical trials can be summarized and counted
.
1.
4 If the clinical trial contains different clinical trial purposes, the clinical trial design shall be carried out separately, including the corresponding sample size estimation
.
For example, the purpose of clinical trials includes: ①Evaluation of the test consistency of in vitro diagnostic reagents and comparison reagents; ②Evaluation of the sensitivity and specificity of in vitro diagnostic reagents used in the differential diagnosis of diseases.
For the above two clinical test purposes, they should be separated Carry out clinical trial design and estimate the minimum sample size
.
1.
5 For quantitative detection reagents, each concentration level within the linear range should have a certain number of sample cases, and focus on fully verifying the detection performance of the medical decision level; for qualitative detection reagents, clinical trial samples should contain a certain number Samples near the positive judgment value (if involved)
.
2.
Using statistical methods to estimate the sample size.
The relevant elements of using statistical formulas to estimate the sample size generally include the expected value of the evaluation index, the acceptance standard of the evaluation index (if applicable), the error probability of type I and type II, and the expected excluding subject off rate
.
The evaluation indicators used for sample size estimation should be specified in the plan
.
The expected value of the evaluation index is set based on the existing clinical data (based on the target population sample), the results of small-sample exploratory trials (if any) or other appropriate evaluation data.
The basis for determining these parameters should be clarified in the clinical trial protocol
.
In general, the type I error probability α is set to two-sided 0.
05 or one-sided 0.
025, and the type II error probability β is set to not more than 0.
2
.
The appendix of this article illustrates several common sample size estimation methods for reference
.
3.
It should be noted that the sample size estimation is only based on the minimum sample size estimation required by statistics.
The sample size of clinical trials should be fully evaluated for clinical performance as the goal to ensure that the various characteristics of the target population have a sufficient number of representative subjects Participants are included in the group, so that the clinical trial results can fully and truly reflect the situation of the target population
.
(7) Control of Bias Bias refers to the systematic errors caused by relevant influencing factors during the design, implementation and result analysis of clinical trials, which cause the evaluation of the safety and effectiveness of the test in vitro diagnostic reagents to deviate from the true value
.
Biases interfere with clinical trials to reach correct conclusions, and it is necessary to prevent them from happening during the whole process of clinical trials
.
1.
Blind method: In the comparative research test of in vitro diagnostic reagents, the test operators and result assessors of the test in vitro diagnostic reagents and the comparison method (if any) should be blinded, so that they do not know the disease diagnosis of the subject during the test.
Or other relevant test results and other information, so as to avoid introducing bias
.
2.
The test of in vitro diagnostic reagents should be performed synchronously with the judgment of clinical reference standards or comparison methods as far as possible to avoid the deviation of the clinical trial conclusions from the true value due to different disease processes or large differences in sample storage time
.
3.
Bias control needs to be carried out at all stages of clinical trial institution selection, subject selection, trial process, and statistical analysis
.
For example: the subject population should represent the characteristics of the target population as much as possible to avoid selection bias; different clinical trial institutions should unify test operations and interpretation standards in clinical trials to avoid the occurrence of central effects
.
(8) Other factors that need to be considered in the design of clinical trials.
Some inherent characteristics of in vitro diagnostic reagents may affect key elements in the design of clinical trials.
When designing clinical trials, they should be considered, including the principles and use of in vitro diagnostic reagents.
the method (including requirements for skill level, etc.
), usage conditions (usage environment and the user), the influence of human factors such as the use
.
For example: 1.
In vitro diagnostic reagents that are expected to be used by consumers themselves.
In clinical trials, in addition to evaluating the clinical performance of the in vitro diagnostic reagents, consumers with no medical background also need to evaluate the cognitive ability of the product instructions and prove Consistency of test results between consumers without medical background and professional inspectors
.
2.
If the testing operation, data processing, and interpretation of in vitro diagnostic reagents have special skills requirements for users, the skill level and training requirements of test personnel in clinical trials should be able to represent the expected users of the product after it is marketed.
Request
.
IV.
Clinical trial quality management In vitro diagnostic reagent clinical trials should meet the relevant requirements of the "Medical Device Clinical Trial Quality Management Practice", safeguard the rights and safety of subjects in the clinical trial process, and ensure that the clinical trial process is standardized, and the results are true, accurate and complete And traceable
.
Clinical trials should cover the whole process of quality management of clinical trials, including clinical trial design, implementation, audit, inspection, inspection, data collection, recording, preservation, analysis, review and reporting
.
The following is an explanation of the issues that require special attention in the quality management of in vitro diagnostic reagent clinical trials: (1) Pre-clinical trial management 1.
Before clinical trials, ensure that the product design has been finalized, and complete the analytical performance evaluation and positive judgment of the in vitro diagnostic reagents.
Value or reference interval research, quality inspection, risk-benefit analysis, etc.
, and the results should be able to support the clinical trial
.
2.
The in vitro diagnostic reagents used in clinical trials are produced in accordance with the relevant requirements of the medical device production quality management regulations and are of qualified quality
.
(2) In principle, clinical trials should be carried out by various clinical trial institutions as far as possible.
If there are large differences in time stages, there should be reasonable explanations to confirm that the same clinical trial protocol is used, and bias and central effects should be applied.
Analysis
.
(3) Data and records 1.
In clinical trials, the principal investigator should ensure that any observations and findings are accurately and completely recorded
.
The source data of clinical trials should include at least: 1.
1 Information about the reagents and instruments used, including name, specification/model, batch number/serial number, quantity, date of receipt, storage conditions, usage, and handling of remaining reagents, etc. .
1.
2 Subject screening and selection records, basic subject information (such as gender, age, enrollment time, etc.
), clinical diagnosis and treatment information, sample inspection records, and adverse events
.
1.
3 Complete records of the source, number, preservation, use, retention, and destruction of samples used in clinical trials
.
1.
4 Signature and date of recorder
.
2.
Clinical trial source data should not be changed at will; when it is really necessary to make changes, the reasons should be explained, signed and dated
.
3.
Sample management and traceability: clinical trial samples should be provided by the clinical trial institution that conducted the trial, should have a unique traceability number, and each sample should be traceable to a unique subject (if there are special circumstances, it should be in the plan and report Described in)
.
4.
The clinical trial data should be traceable, and the data in the clinical trial report, case report form, clinical trial data sheet and other documents should be consistent and traceable to the source file
.
(4) Management of reagents required for clinical trials.
The transportation, use, and storage of in vitro diagnostic reagents, comparison reagents and other supporting reagents (such as nucleic acid extraction reagents, etc.
) and instruments and equipment used in clinical trials shall comply with Related requirements
.
V.
Others 1.
For specific in vitro diagnostic reagents, specific clinical test methods, statistical methods, sample size estimation, etc.
may have specific requirements, and the sponsor should make scientific selection and design based on specific conditions
.
If there are technical review guidelines for related products, the relevant requirements should be referred to
.
2.
Some clinical trials use in vitro diagnostic reagents and nucleic acid sequence determination, GC-MS/MS and other laboratory testing reference methods for comparative research.
These methods are not routine clinical testing techniques and require specialized equipment and test conditions.
Some clinical trial institutions do not have relevant testing conditions
.
In such cases, applicants should choose clinical trial institutions with corresponding conditions to carry out clinical trials as much as possible.
Some clinical trial institutions that do not have the testing conditions can entrust this part of the test to a specialized sequencing agency or a trial with certain testing qualifications.
The laboratory conducts testing and approves the test results
.
Research data and quality control data and other product registration application should be submitted clinical trial institution entrusted with the organization of the commission documents, comparative method approach
.
However, the sponsor’s laboratory shall not be entrusted to carry out relevant tests
.
The test results directly commissioned by the sponsor shall not be submitted as clinical trial data
.
Attachment: Examples of sample size estimation methods The sponsor should select an appropriate sample size estimation method based on the specific characteristics of the product, statistical analysis models and other factors, and specify the basis for determining the sample size in the plan
.
At the same time, the possible drop-off rejection rate should be fully considered, and the sample size requirement should be set reasonably
.
Here are examples of several sample size estimation methods
.
1.
Examples of sample size estimation for qualitative testing 1.
When the evaluation index has a definite clinical acceptance standard, it is necessary to prove that the product evaluation index meets the requirements of the acceptance standard
.
At this time, the single-group target value method sample size formula can be used to estimate the minimum sample size
.
In the formula, n is the sample size; Z1-α/2, Z1-β are the scores of the standard normal distribution of significance level and power, P0 is the clinical acceptance standard of the evaluation index, and PT is the evaluation of the test in vitro diagnostic reagent The expected value of the indicator
.
Generally, in the clinical trials of in-vitro diagnostic reagents, comparative trials with similar products on the market can set appropriate clinical acceptance standards according to clinical needs, and use the above formula to estimate the minimum sample size
.
Example: The qualitative detection reagent clinical test adopts the method of comparing the test in vitro diagnostic reagents with similar products on the market.
According to the clinical needs, the positive and negative coincidence rates should reach 85% and 90% respectively.
According to the results of the exploratory test, the test in vitro The positive and negative coincidence rates of diagnostic reagents and contrast reagents are expected to reach 90% and 94%, respectively
.
The minimum sample sizes of the positive group (n+) and negative group (n-) are estimated as follows: According to the above estimation, the total number of sample cases is estimated to be 750
.
According to the dropout rate of 10%, at least 834 subjects should be enrolled.
In the actual enrolled population, the number of samples in the positive and negative groups should meet the minimum requirements of n+ and n-, respectively
.
It should be noted that in addition to meeting the minimum sample size requirements of the above statistical estimation, the sample size of clinical trials should also ensure that the included cases cover various characteristics of the subjects; if different sample types are involved, cases of different sample types should also be considered.
number of requirements
.
2.
In the parameter estimation of clinical trials, only the evaluation index is guaranteed to meet the expected accuracy level (the width of the confidence interval is constant), and the clinical acceptance standard is not set, the following formula can be used: In the formula, n is the sample size, Z1- α / 2 is the quantile confidence standard normal distribution, P is the expected value of the evaluation index, Δ P is the size of the allowable error
.
It should be noted that the values ​​of P and Δ should have sufficient basis.
Unless there are special reasons, it is not recommended to set Δ>0.
05.
When the expected value is higher, better accuracy should be considered
.
Using the above formula, the sample size of subjects with the target disease state (positive) or subjects without the target disease state (negative) can be estimated based on the expected value of sensitivity or specificity
.
For example: a certain marker detection reagent is used in the auxiliary diagnosis of related diseases.
Through the analysis of existing data, it is known that the sensitivity of the detection reagent is expected to be 85%, and the specificity is expected to be 90%.
Clinical trials use in vitro diagnostic reagents.
Method of comparative research with clinical reference standards to evaluate the clinical performance of test in vitro diagnostic reagents
.
If the allowable error Δ is 0.
05, the minimum sample size (n+) of subjects with the target disease state (positive) is estimated as: The minimum sample size (n-) of subjects without the target disease state (negative) is estimated as: Based on the above estimation, the total number of sample cases is estimated to be 334 cases
.
According to the dropout rate of 10%, at least 371 subjects should be enrolled.
Among the subjects actually enrolled, subjects with the target disease state (positive) and subjects without the target disease state (negative) The number of sample cases should meet the minimum requirements of n+ and n- above
.
In this clinical trial, in addition to meeting the minimum sample size requirements, subjects with target disease states (positive) should also ensure that cases of various types of diseases, different disease processes, and different severity of diseases are covered; they do not have the target disease Subjects with status (negative) need to cover various subject samples required for clinical specificity evaluation
.
If the expected sensitivity or specificity of different subgroups is different, it may be necessary to stratify into groups and estimate the sample size of the subgroups separately
.
3.
It should be noted that when the evaluation index P is close to 100%, the above two sample size estimation methods may not be applicable.
You should consider choosing a more appropriate method for sample size estimation and statistical analysis, such as accurate probability method
.
2.
Estimation of sample size for quantitative testing For quantitative testing, appropriate statistical methods can also be selected to estimate the sample size for appropriate evaluation indicators
.
In the clinical trial protocol, it is recommended that the clinical acceptance criteria corresponding to the selected evaluation indicators should be given at the same time, and the basis for determination should be provided
.