IN FORCE undated

52 Inviting comments on standard IVD evaluation protocols drafted by ICMR and CDSCO 2025-Aug-26 5218 KB

Document text

ICMR-CDSCO/IVD/TB/PROTOCOLS/2025 Page 1 of 1

Indian Council of Medical Research and Central Drugs Standard Control Organization Department of Health Research and Drugs Controller General of India Ministry of Health and Family Welfare Government of India Document No.: ICMR-CDSCO/IVD/TB/PROTOCOLS/2025

Subject: Inviting comments on standard IVD evaluation protocols drafted by ICMR and CDSCO

Licensure of In-Vitro Diagnostics (IVDs) under Medical Devices Rules 2017 requires a detailed evaluation protocol for the performance evaluation of IVDs to evaluate their quality and performance. To facilitate this process, the Indian Council of Medical Research (ICMR) and CDSCO have come together to draft standard evaluation protocols for use by IVD manufacturers testing labs in India. Currently, the following protocols for Tuberculosis have been developed by ICMR and CDSCO:

Analytical Performance Evaluation of In-Vitro Diagnostics for Pulmonary Tuberculosis
Clinical Performance Evaluation of In-Vitro Diagnostics for Pulmonary Tuberculosis
Clinical Performance Evaluation of In-Vitro Diagnostics for Pulmonary Drug Resistant Tuberculosis

These protocols are now being placed in the public domain for comments from relevant stakeholders. This window of opportunity will close on 7th September 2025, and, once finalized, there will be minimal scope for change in these documents. Therefore, all interested stakeholders are requested to provide their comments before 7th September 2025, at ivdevaluation@gmail.com as per the enclosed format. Once the public consultation period concludes, all comments will be reviewed and considered in finalizing the draft protocols before final clearance by ICMR and CDSCO.

Dated: 27th August 2025
Place: New Delhi

STANDARD IVD PERFORMANCE EVALUATION PROTOCOL STAKEHOLDER FEEDBACK FORM

S.N. Name of the Protocol Document No. Page No. Line No. Current Text Proposed Text Explanation/Reference

Name:

Designation and Affiliation:

ICMR-CDSCO STANDARD PERFORMANCE EVALUATION PROTOCOLS

5 ANALYTICAL PERFORMANCE EVALUATION OF 6 IN-VITRO DIAGNOSTICS FOR PULMONARY 7 TUBERCULOSIS 8 ICMR-CDSCO/IVD/TB/PROTOCOLS/1/2025 9

10 11 DIVISION OF COMMUNICABLE DISEASES, ICMR IN VITRO DIAGNOSTICS DIVISION, CDSCO AUGUST, 2025 India

Analytical Performance Evaluation of IVD for Pulmonary Tuberculosis

2 | P a g e

12 LIST OF CONTRIBUTORS 13

15 A. Working Group: 16

Dr Sivakumar, Scientist E and Head, Department of Bacteriology, ICMR-NIRT, 17 Chennai 18
Dr Madhumathi J, Scientist D, Division of Communicable Diseases (CD), ICMR, 19 New Delhi 20

22 B. Review Committee: 23

Dr Amita Jain, Professor and Head, Department of Microbiology, KGMU 25
Dr RM Pandey, ICMR- Dr A.S. Paintal Distinguished Scientist Chair 26
Dr. Camilla Rodrigues, Senior Consultant, P.D. Hinduja Hospital, Mumbai 27
Dr Gita Nataraj, Professor Emeritus, Microbiology, Seth GS Medical College and 28 KEM Hospital, Mumbai
29
Dr Ashutosh Aggarwal, Professor and Head, Pulmonary Medicine, PGIMER 30
Dr Venkataraghava Mohan, Professor & Head, Dept. of Community Health and 31 Development, CMC Vellore 32
Mr. Pramod Meshram, Deputy Drugs Controller, Central Drugs Standard Control 33 Organization, Directorate General of Health Services, Ministry of Health and Family 34 Welfare, Government of India 35
Dr. Sella Senthil, Assistant Drugs Controller, Central Drugs Standard Control 36 Organization, Directorate General of Health Services, Ministry of Health and Family 37 38
Dr. Nivedita Gupta, Scientist-G and Head of the Division of Communicable Diseases, 39 ICMR Headquarters, Department of Health Research, Ministry of Health and Family 40 41

43 44

3 | P a g e

46 47

48 Table of Contents 49

50 S.No Content Page Number I. Background 4 II. Purpose 4 III. Study Design 4 IV. Ethical Considerations 4 V. Blinding of Laboratory Staff 5 VI. Procedure i. Preparation of Evaluation sites/laboratories ii. Exclusion iii. Reference tests iv. Preparation of samples v. Reference Strains vi. Sample size and sample panel composition vii. Limit of Detection (LOD) Assay viii. Reproducibility ix. Inclusivity and exclusivity x. Cross-contamination/carry-over xi. Resolution of discrepancy

6 6 6 6 7 7 10 13 13 14 14 VII. Statistical Analysis Plan 15 VIII. Acceptance Criteria 15 IX. Publication Rights 16 X. Important Note 16 XI. References 17 XII. Performance Evaluation Report Format 18

4 | P a g e

51 I. Background 52 CDSCO and ICMR, New Delhi, have aimed at facilitating the availability of Quality-Assured 53 diagnostic kits appropriate for use in India. This protocol gives the methods to be used for 54 evaluating the analytical performance characteristics of the in-vitro diagnostic test in detecting 55 pulmonary tuberculosis and drug-resistant tuberculosis. 56

57 II. Purpose 58 To evaluate the performance characteristics of nucleic acid amplification tests (NAAT) forthe 59 diagnosis of pulmonary Mycobacterium tuberculosis (MTB)using irreversibly de-identified 60 leftover archivedor spiked sputum samples. 61

62 III. Study Design 63 Analytical validation of IVD using irreversibly de-identified leftover clinical/spiked samples. 64

65 IV. Ethical Considerations 66

Leftover sputum specimens collected for routine diagnostic evaluation from patients who 67 are suspected of having TB shall be used. No additional specimens should be requested.
68
The probability of harm or discomfort anticipated in the research is nil or not expected. 69
Performance evaluation activities using irreversibly de-identified leftover clinical 70 samples are exempt from ethics approval as per ICMR’s Guidance on Ethical 71 Requirements for Laboratory Validation Testing, 2024.
72
Investigators are required to submit a self-declaration form, as outlined in the ICMR 73 guidelines, to the institutional authorities and ethics committee for information. 74
The protection of privacy of participants should be ensured by using de-identified 75 samples and encrypting the patient identifiers. 76
Respect for the dignity of participants shall be prioritized. 77

78 79

5 | P a g e

80 V. Blinding of Laboratory Staff 81 To ensure the rigor of the evaluation process, laboratory staff performing the evaluation should 82 be blinded to the status of the clinical samples. The PI of the evaluation exercise should remain 83 unblinded, i.e., privy to the status of the samples. Another senior laboratory staff member 84 selected by the PI may remain unblinded and carry out coding of samples and dispensing them 85 into similar-looking vials to be used for testing, and maintain the database of results. Staff 86 performing the reference test and the test under evaluation, interpretation of the test result, and 87 entering the results against the coded samples in the database, should remain blinded to the status 88 of samples till the completion of evaluation. The data should be analyzed only by the PI of the 89 evaluating lab. Refer to Fig. 1. 90

94 Figure 1 Blinding in evaluation exercise 95

96 97

6 | P a g e

98 VI. Procedure 99

100

Preparation of Evaluation sites/laboratories 101 A. The laboratory must be approved by the National TB Elimination Program (NTEP). 102 B. Accreditation for at least one Quality management system [accreditation for Testing Lab / 103 Calibration Lab (ISO/IES 17025), Medical Lab (ISO 15189), PT provider ISO/IEC 104 17043 or CDSCO approved Reference laboratory]. 105
Exclusion 106  Extra-pulmonary samples 107  Specimens with > 1 freeze-thaw cycle (or according to IFU, if specified) 108  Any exclusion criteria stated in the product IFU 109

110 3. Reference tests 111  For detection of MTB: Mycobacterium Growth Indicator Tubes (MGIT) liquid culture. 112  For MDR-TB: MGIT drug sensitivity testing (DST) 113

114 4. Preparation of samples 115  For LOD studies - MTBC-negative sputum: smear-negative and NAAT-negative 116 sputum should be used for the spiking analytic studies 117  For analytical sensitivity and specificity: Well characterized archived samples (sputum 118 or processed sputum); MTB positives, MTB negatives and Non-Mycobacterium 119 tuberculosis (NTM) samples confirmed by liquid MGIT culture
120  For drug sensitivity:MTB and NTM clinical isolatesthoroughly characterized through 121 MGIT DST and sequencingshould be used.
122  For inclusivity/exclusivity, resistance detection, and cross-contamination, mycobacterial 123 strains should be diluted into 7H9 medium at the required concentrations.
124  The concentrations (cfu/mL) should be estimated by adjusting the bacterial suspension 125 density to the McFarland standards. 126

127

7 | P a g e

128 5. Reference Strains 129 The National Institute for Biological Standards and Control (NIBSC) internal reference standard for 130 Mycobacterium tuberculosis (H37Rv) DNA for Nucleic Acid Amplification Test (NAAT) based 131 assays (NIBSC code: 20/152) will be used for the LOD assay.It was established as the 1stWHO 132 International Standard for Mycobacterium tuberculosis (H37Rv) DNA for NAAT-based assays in 133 2021. The intended uses of this material are for calibration of secondary or in-house reference 134 materials used in the assays for the molecular detection of M. tuberculosis DNA.It may also be used 135 for assay validation and monitoring the limit of detection of rapid diagnostic tests. This preparation 136 contains an arbitrary unitage of 6.3 log10 (or 2 million) IU per vial.
137

138 6. Sample size and sample panel composition 139 With an anticipated sensitivity of 90% and relative precision of 7%, a minimum of 87 140 confirmed MTB positive samples by MGIT culture will be required for testing analytical 141 sensitivity. With an anticipated specificity of 95% with 5% relative precision, the minimum 142 sample size required for analytical specificity is 81 confirmed MTB negative samples by 143 MGIT culture. To rule out NTM detection, with an assumed sensitivity of 90% and relative 144 precision of 10%, around 50 confirmed NTM samples may be included to evaluate the index 145 test kit. Hence, approximately 100 confirmed MTB positives, 100 confirmed MTB negatives 146 and 50 NTM samples will be used for pre-validation studies. 147

148 The proposed evaluation study will be done using Sputum/MTB isolates stored at the biobank 149 facility of the National TB reference laboratories (NRLs) or the pre-validation labs. The 150 stored sputum/MTB isolate/processed sample/DNA samples will be of the following 151 categories and sub-categories. 152 Category 1: Positive for MTB by MGIT culture (N = 100) 153 Category 2: Negative for MTB by MGIT culture (N = 150)
154 Within the MTB negative group, we propose the following two sub-categories: 155 i. Negative for all Mycobacteria (N = 100) 156 ii. Positive for Non-Tuberculous Mycobacterium (N = 50) 157

158

8 | P a g e

Category 3: If resistance detection has to be carried out, within the MTB positive group, we 159 propose to use the following sub-categories: 160 i. Sensitive to Rifampicin and Isoniazid, individually and combined (N = 100) confirmed 161 by Drug susceptibility testing on MGIT liquid culture. 162 ii. Resistance to both Rifampicin and Isoniazid (N = 100) as detected by Drug susceptibility 163 testing on MGIT liquid culture. 164 iii. Isoniazid mono-resistance (N = 45) as detected by DST on MGIT liquid culture. 165 iv. Fluroquinolone resistance (N=45) (If applicable for the index test) as confirmed by DST 166 on MGIT liquid culture. 167

168 Table 1: Sample size calculation with 95% confidence level 169

170 Anticipated Sensitivity Relative Precision Sample size 90% 5% 171 90% 10% 43 90% 7% 87 95% 5% 81 95% 10% 20 95% 7% 41

171 172

9 | P a g e

Analytical sensitivity and specificity: 173

174 175

176

177

178 Figure 2. Flowchart for Analytical Validation for detection of; a) MTB detection, b) 179 MDR-TB
180

181

182

183

184

185

186 a) b)

10 | P a g e

Limit of Detection (LOD) Assay 187 The 95% LOD is defined as the minimum concentration of bacterium, expressed as CFU/ml or 188 genomic DNA copy numbers/mL, in a sample volume that can be detected in 95% of tests. 189 Finalize the LOD at least one concentration with a hit rate above 95% and two concentrations 190 with hit rates between 10% and 90%.LOD should be always done with NIBSC H37Rv (20/152) 191 standard and only reported in IU/ml or CFU/ml 192

193 Preparation of samples for LOD 194

The volume of sputum required for LOD is based on the IFU (Instruction for use) from the 195 index test manufacturer, which generally varies between 1-2 ml of sputum. 196
A minimum of 200ml of NAAT negative sputum is required for the full LOD studies for a 197 single index test. 198
Sputum samples which are negative by Smear and GeneXpert will be stored at -20C and once 199 the required amount is obtained the samples will be pooled and tested for MTB using 200 molecular and phenotypic test to prove no growth of MTB in the pooled samples. 201
To perform the assay it may take two weeks to one month based on the multiplicities of test 202 suggested in the IFU after the required volume of sputum is collected. 203 Spiking of sputum samples 204
The spiked sputum will be used to determine the LOD of the test kit. About 1.8 ml of negative 205 sputum specimen will be spiked with 200 ul of the respective diluted suspension series of M. 206 tuberculosis H37Rv. 207
These dilutions will be added to the sputum to get the final concentration (10000, 1000, 100, 208 and 10 IU/ml). Before spiking, the culture for CFU will be set up for the different dilutions.
209
NIBSC reference standard will be reconstituted as directed by NIBSC using 1 mL nuclease 210 free molecular biology grade purified water (MBGPW). From this stock 100 µL will be 211 diluted ½ to get 10,00,000 IU/ml and serially diluted to give 100000, 10000, 1000 and 100 212 IU/ml with MBGW.
213
Each dilution of the WHO International Standard, will be tested 24 times. The 24 replicates 214 will be performed over at least three days by at least two users and, for low-throughput 215 instruments, on at least three different instruments, or sets of instruments if applicable (e.g., 216

11 | P a g e

DNA preparation and amplification instruments). For low through-put instruments, the 217 number of testing days may be increased. 218 5. Each lot shall comprise different production (or manufacturing, purification, etc.) runs of 219 critical reagents. Inter-lot variation must be evaluated by appropriate statistical means. 220 6. Lowest dilution at which the test detects M.tb will be determined a LOD, the corresponding 221 CFU will also be counted and reported in terms of CFU/per ml. The LOD will be presented as 222 IU/mL for each dilution. 223 7. Analytical sensitivity shall be estimated by determining the 95% LOD with 95% confidence 224 intervals (e.g., by probit analysis).
225 8. If there are more than four invalid results with the same specimen (i.e. dilution) overall, then 226 the specimen will be retested to get at least 20 valid results for each dilution. For tests that 227 include a claim for drug resistance testing, at least 20 valid results (i.e., sensitive or resistant) 228 for each of the claimed drugs should be obtained for each dilution. 229 9. To arrive at the LOD a probit analysis should be performed, Probit analysis is defined as a 230 specialized form of regression analysis applied to binomial response variables, transforming a 231 concentration-response curve into a straight line for analysis through methods like least 232 squares or maximum likelihood regression. It is primarily used in molecular biology 233 measurement procedures, such as PCR, to determine the detection probability of analytes at 234 various concentrations.
235 LOD for detection of drug resistance 236

To test the drug resistant MTB strains, well-characterized MTBC strains of known 237 concentration (expressed as CFU/mL) shall be spiked into each claimed MTBC negative 238 specimen type. DR strains shall be characterized by sequencing. 239
Relevant DR strains (as mentioned in table below) shall be spiked into each claimed MTBC- 240 negative specimen type (e.g., raw and/or processed sputum, and each claimed extra- 241 pulmonary specimen).
242
If the assay detects resistance to more than 1 target drug, the LOD for each target drug in 243 addition to a composite resistance LOD, defined as the highest LOD among the tested target, 244 shall be reported. 245

12 | P a g e

Analytical sensitivity for resistance detection shall be estimated as the lowest number of 246 colony-forming units (CFU) per specimen that can be reproducibly distinguished from 247 negative specimens with 95% confidence.
248
The determination shall comprise 24 replicate tests (8 replicate tests on each of 3 days) of a 249 minimum 8 8-member 0.5log10 dilution panel. The replicate testing shall be conducted on 250 three different days using 2 lots, and at least 2 dilution series shall be tested. 251 Table 2: Anti-mycobacterial drugs and common mutations 252 S.No Drugs Resistance mutation of strains to be tested
1 Isoniazid katG_S315T and fabG1_c-15t 2 Rifampicin rpoB_S450L; rpoB_D435V; rpoB_H445Y; rpoB_H445D; rpoB_D435Y; rpoB_S450W; rpoB_L452P; rpoB_H445L; rpoB_S450F; rpoB_L430P; rpoB_H445R; one rpoC mutation 3 Levofloxacin (CC) LFX2,3
gyrA_A90V, gyrA_D94G, gyrA_D94H, gyrA_D94N, gyrA_D94Y, gyrA_S91P 4 Moxifloxacin (CC and CB) gyrA_A90V, gyrA_D94G, gyrA_D94H, gyrA_D94N, gyrA_D94Y, gyrA_S91P 5 Bedaquiline Rv0678_LoF, pepQ_LoF, atpE_p.Ala63Pro 6 Linezolid rplC_p.Cys154Arg, rrl_n.2814G>T
7 Ethambutol embB_M306L, embB_M306V, embB_Q497R 8 Delamanid ddn_LoF, ddn_p.Leu49Pro, fbiC_LoF 9 Pyrazinamide pncA_ V139A, pncA_ V139G 10 Amikacin rrs_ A1401G, rrs_ A1401G, rrs_G1484T, eis /promoter_ C-12T, eis /promoter_C-14T
11 Kanamycin rrs_ A1401G, rrs_ A1401G, rrs_G1484T, eis /promoter_ C-12T, eis /promoter_C-14T 12 Capreomycin rrs_ A1401G, rrs_ A1401G, rrs_G1484T, eis /promoter_ C-12T, eis /promoter_C-14T 13 Ethionamide fabG1_c-15t, inhA_S94A, fabG1_ T-8C 14 Pretomanid# ddn_LoF, ddn_p.Leu49Pro, fbiC_LoF 15 Cycloserine Alr_C-8T, alr_M319T, alr_Y364D, ald_T-32C, ddlA T365A 16 PAS thyA T22A, folC I43T, folC R49W

13 | P a g e

Reproducibility 253 Within-run (same operator, same measuring system, same operating conditions, and same 254 location), Between-run, -lot, -day, -site, -operator.
255
Three specimens will be used; MTB sensitive (H37Rv), MTB resistant and MTB negative. 256
The effect of operator-to-operator variation on IVD performance will be included as part of 257 the precision studies. Each lot will comprise different production (or manufacturing, 258 purification, etc.) runs of critical reagents. 259
The nucleic extraction/purification component will also be considered for estimating 260 precision.
261
Contrived specimens will be used (i.e., MTBC strains with specific/most common mutations 262 in the target genes spiked into a clinical matrix claimed in the IFU) for repeatability and 263 reproducibility studies. 264
DR specimens at the concentrations specified for each DRTB (i.e. RR-TB, Hr-TB, MDR-TB, 265 TB resistant to fluoroquinolones) as described in the table on resistance detection.
266
If there are two or more invalid results for the same specimen in the same run, then the run 267 should be repeated for this specimen. Invalid results should be reported. 268
Results will be statistically analyzed by ANOVA or other methods to identify and isolate the 269 sources and extent of any variance.
270
Furthermore, the percentage of correctly identified, incorrectly identified, and invalid results 271 will be compiled for each specimen and separately categorized by site, lot, and other factors. 272

273 9. Inclusivity and exclusivity 274

Inclusivity MTBC stains: M. bovis 275
Exclusivity NTM strains: M. avium, M. kansasii, M. intracellulare 276
Representative MTBC and non-tuberculosis mycobacteria (NTM) strains will be tested in 277 triplicate forinclusivity and exclusivity verification. 278
Resistance detection: For assays with a claim for detection of drug resistance, the 279 applicable specimens from the resistancedetection panelwill be tested in triplicate. 280
The concentration of MTBC isolates used in inclusivity studies will be at levels at or near 281 the specific LOD and will be confirmed by plating/ counting bacterial CFUs (estimated 282

14 | P a g e

using Truenat). 283 6. The selection of specific MTBC strains with relevant genetic variations linked to DR will 284 be made to support the claims in the IFU.
285 7. This will involve testing strains that carry the most common mutations, including 286 associated or interim resistance mutations, covering at least 80% of the resistance 287 mechanisms observed globally for each of the assay target drugs (as shown in table 2).
288

289 10. Cross-contamination/carry-over 290

The experiment will allow the determination of the well-to-well or vial-to-vial cross- 291 contamination rate ofhigh-throughput platforms or potential carryover in low-throughput 292 instruments.
293
This will be assessed byalternating one high-positivespecimen with one negative 294 specimen and repeating thissequence twenty times.
295
For high-throughput assays, this will be achieved by alternating high-positive andhigh- 296 negative specimens in the same plate/run.
297
For low-throughput assays, each sequence of highly positivespecimens followed by 298 negative specimens should be done on the same instrument.
299
If more than one instrument isused, each run (i.e same instrument and same day) should 300 include a minimum of 2 sets of alternating high-positive and negative specimens. 301
Contrived specimens prepared by spiking MTBC strains into MTBC negative clinical 302 sputum will be used for these studies. 303

304

305

306

307

308 Note: The strains used for assessment of reproducibility, inclusivity/exclusivity, resistance detection, and carry-over may be commercially acquired or locally prepared, well-characterized strains (by phenotypic DST and sequencing).

15 | P a g e

Resolution of discrepancy: 309

310  The results of MGIT culture should be used to resolve any discrepancy in detection of 311 MTB 312  Results of phenotypic DST and sequencing should be used to resolve discrepancy in 313 detection of MDR-TB. 314

315 VII. Statistical Analysis Plan 316

The index molecular test should be evaluated for its analytical sensitivity and analytical 317 specificity.
318
95% Confidence interval should be calculated for each of the parameters. 319 % Sensitivity = Positives by index test

x 100 = [a/a+c] * 100 320

         Confirmed positives by MGIT culture

321

322 % Specificity= Negatives by index test

x 100 = [d/b+d] * 100 323 Confirmed negatives by MGIT culture
324

325

326 VIII. Acceptance Criteria 327 Acceptance criteria for Diagnostic tests:
328 Expected sensitivity: ≥90% 329 Expected specificity: ≥95% 330 Sample Size: ~ 100 confirmed MTB positives (by MGIT culture), ~ 100 confirmed MTB 331 negatives (by MGIT culture) and ~ 50 NTM samples (confirmed by culture and identification) 332 333

16 | P a g e

Acceptance criteria for Screening tests:
334 Test Type Minimal Accuracy Optimal accuracy High Sensitivity high specificity screening test 90% Sensitivity
80% specificity 95% Sensitivity
95% specificity High Sensitivity screening test 90% Sensitivity
60% specificity 95% Sensitivity
85% specificity High specificity screening test 60% Sensitivity
98% specificity 70% Sensitivity
98% specificity Source: WHO TPP 2025 335

336 IX. Publication Rights 337 The PI(s) of the evaluating labs shall retain publication rights of the evaluation as lead author(s). 338

339 IMPORTANT NOTE 340 Once a kit is determined to be “Not of Standard Quality”, following the procedure outlined 341 in this document, no further requests for repeat testing of that kit will be accepted.
342

343 Any request of re-validation from the same manufacturer for the same test type will only 344 be entertained after a minimum of 3 months and only if a high-level technical summary of 345 modifications or functional improvements to the kit design is submitted, without explicit 346 disclosure of proprietary information. 347

348 Clinical samples are precious, therefore, repeat evaluation of a kit using the same/ different 349 well-characterized sample panel at a different laboratory may be considered only for kits 350 which claim high performance characteristics (sensitivity and specificity 95% and above), 351 but which fail the performance evaluation by a margin of 5%. 352

353 354

17 | P a g e

References 355

Technical Specifications Series for Submission to WHO Prequalification – Diagnostic 356 Assessment. TSS 17 In vitro diagnostic medical devices used for the qualitative detection of 357 Mycobacterium tuberculosis complex DNA and mutations associated with drug-resistant 358 tuberculosis.https://iris.who.int/bitstream/handle/10665/366068/9789240055865-eng.pdf. 359
Miotto, P., Tessema, B., Tagliani, E., Chindelevitch, L., Starks, A. M., Emerson, C., Hanna, 360 D., Kim, P. S., Liwski, R., Zignol, M., Gilpin, C., Niemann, S., Denkinger, C. M., Fleming, J., 361 Warren, R. M., Crook, D., Posey, J., Gagneux, S., Hoffner, S., Rodrigues, C., … Rodwell, T. 362 C. (2017). A standardised method for interpreting the association between mutations and 363 phenotypic drug resistance in Mycobacterium tuberculosis. The European respiratory 364 journal, 50(6), 1701354. https://doi.org/10.1183/13993003.01354-2017. 365
Indian Catalogue of Mycobacterium tuberculosis Mutations and their Association with Drug 366 Resistance, Version 2.0 2024, ICMR-NIRT. 367 https://www.nirt.res.in/pdf/mutation_catalogue_v2.pdf. 368
Target product profiles for tuberculosis screening tests. Geneva: World Health Organization; 2025. 369 Licence: CC BY-NC-SA 3.0 IGO. Available at: 370 https://iris.who.int/bitstream/handle/10665/382179/9789240113572-eng.pdf?sequence=1.
371
WHO manual for the preparation of secondary reference materials for in vitro diagnostic assays 372 designed for infectious disease nucleic acid or antigen detection: calibration to WHO International 373 Standards. Available at: https://www.who.int/publications/m/item/annex-6-trs-no-1004 374

375

376

377 378

18 | P a g e

PERFORMANCE EVALUATION REPORT FORMAT 379 Performance Evaluation Report For MTB/MDR-TB Kit 380

381 Name of the product (Brand/generic)

Name and address of the legal manufacturer

Name and address of the actual manufacturing site

Name and address of the Importer

Name of supplier: Manufacturer/Importer/Port office of CDSCO/State licensing Authority

Lot No /Batch No.:

Product Reference No/ Catalogue No

Type of Assay

Kit components

Manufacturing Date

Expiry Date

Pack size (Number of tests per kit)

Intended Use

Number of Tests Received

Regulatory Approval: Import license / Manufacturing license/ Test license

License Number: Issue date: Valid Upto:

Application No.

Sample Panel Sample type

Positive samples (provide details: strong, moderate, weak)

Negative samples (provide detail: clinical/spiked, including cross reactivity panel)

382 Results:
383

384

Reference assay ……….…………… (MGIT/MGIT DST for RIF/INH/FQ/others)

Positive Negative Total Name of MTB or MDR-TB kit Positive

Negative

Total

385

Estimate (%) 95% CI Sensitivity

Specificity

386

387

19 | P a g e

388 Conclusions: 389

390 o Sensitivity, Specificity 391 o Performance: Satisfactory / Not satisfactory 392

393 (Sensitivity and specificity have been assessed in controlled lab setting using kits provided by the manufacturer from 394 the batch mentioned above using ….. sample. Results should not be extrapolated to other sample types.) 395

396 DISCLAIMERS 397

398

This validation process does not approve / disapprove the kit design 399
This validation process does not certify user friendliness of the kit / assay 400

401 Note: This report is exclusively for ………………………Kit (Lot No……), version …………with the 402 gene targets …………………...manufactured by …………… (Supplied by ……….). 403

404 Evaluation Done on …………………… 405

406 Evaluation Done by …………………………. 407

408

409

410

411 Signature of Director/ Director-In-charge …………………… Seal ………………………… 412

413 ****End of the Report 414

415

416

417

ICMR-CDSCO STANDARD PERFORMANCE EVALUATION PROTOCOLS

5 FIELD PERFORMANCE EVALUATION OF IN-VITRO 6 DIAGNOSTICS FOR PULMONARY DRUG 7 RESISTANT TUBERCULOSIS 8

ICMR-CDSCO/IVD/TB/PROTOCOLS/3/2025

11 DIVISION OF COMMUNICABLE DISEASES, ICMR IN VITRO DIAGNOSTICS DIVISION, CDSCO AUGUST, 2025 India

DRAFT Field Performance Evaluation of IVD for Pulmonary DR-TB

2 | P a g e

13 LIST OF CONTRIBUTORS 14

16 A. Working Group: 17

Dr Sivakumar, Scientist E and Head, Department of Bacteriology, ICMR-NIRT, Chennai 18
Dr Joy Sarojini Michael, Professor, CMC, Vellore 19
Dr Shubhada Shenai, Senior Scientist, FIND 20
Dr Madhumathi J, Scientist D, Division of Communicable Diseases (CD), ICMR, New Delhi 21

23 B. Review Committee: 24

Dr Amita Jain, Professor and Head, Department of Microbiology, KGMU 26
Dr RM Pandey, ICMR- Dr A.S. Paintal Distinguished Scientist Chair 27
Dr. Camilla Rodrigues, Senior Consultant, P.D. Hinduja Hospital, Mumbai 28
Dr Gita Nataraj, Professor Emeritus, Microbiology, Seth GS Medical College and 29 KEM Hospital, Mumbai
30
Dr Ashutosh Aggarwal, Professor and Head, Pulmonary Medicine, PGIMER 31
Dr Venkataraghava Mohan, Professor & Head, Dept. of Community Health and 32 Development, CMC Vellore 33
Mr Pramod Meshram, Deputy Drugs Controller, Central Drugs Standard Control 34 Organization, Directorate General of Health Services, Ministry of Health and Family 35 36
D. Sella Senthil, Assistant Drugs Controller, Central Drugs Standard Control 37 Organization, Directorate General of Health Services, Ministry of Health and Family 38 39
Dr Nivedita Gupta, Scientist-G and Head of the Division of Communicable Diseases, 40 ICMR Headquarters, Department of Health Research, Ministry of Health and Family 41 42

DRAFT

3 | P a g e

47 Table of Contents 48

49 S.No Content Page Number I.
Background 4 II.
Purpose 4 III.
Study Design 4 IV.
Ethical Considerations 4 V.
Blinding of Laboratory Staff 6 VI.
Procedure i. Preparation of Evaluation sites/laboratories ii. Study Participants iii. Eligibility of Participants iv. Reference and Index tests
v. Sample size
vi. Implementation Plan vii. Sample collection, processing and storage viii. Laboratory Tests ix. Data Analysis and resolution of discrepancy x. Quality Control (QC) measures

8 8 8 9 9 11 13 14 14 15 VII.
Statistical Analysis Plan 15 VIII.
Acceptance Criteria 16 Important Note 16

References 16

Performance Evaluation Report Format 17

DRAFT

4 | P a g e

I. Background 52 CDSCO and ICMR, New Delhi, have aimed at facilitating the availability of Quality-Assured 53 diagnostic kits appropriate for use in India. This protocol gives the methods to be used for 54 evaluating the clinical performance characteristics of the in-vitro diagnostic test in detecting 55 pulmonary drug resistant tuberculosis (DR-TB). 56

57 II. Purpose 58 To evaluate the clinical performance characteristics of nucleic acid amplification tests (NAAT) 59 for diagnosis of pulmonary drug resistant tuberculosis (DR-TB) using prospectively collected 60 sputum samples in clinical settings. 61 Primary Objectives 62

To determine the diagnostic accuracy of new multi-drug resistant (MDR) NAAT test against 63 culture based drug sensitivity testing (DST) in detecting first line drug resistance [Rifampicin 64 (RIF), Isoniazid (INH)] among the microbiologically confirmed TB patients (positive by 65 smear or NAAT test).
66
To determine the diagnostic accuracy of new NAAT test against culture-based drug 67 sensitivity testing (DST) in detecting fluroquinolone (FQ) drug resistance among MDR-TB/ 68 RR-TB pulmonary tuberculosis patients
69

70 III. Study Design 71 Cross-sectional prospective multi-centric diagnostic accuracy study of IVD for detection of 72 pulmonary drug resistant TB, using Mycobacterium Growth Indicator Tube culture and drug 73 sensitivity testing (MGIT-DST) as the microbiological reference standard. 74

75 IV. Ethical Considerations 76

The study should be compliant to the ICMR’s Guidance on Ethical Requirements for 77 Laboratory Validation Testing, 2024. Performance evaluation activities using irreversibly de- 78 identified leftover clinical samples are exempt from ethics approval as per ICMR’s Guidance 79 on Ethical Requirements for Laboratory Validation Testing, 2024. Investigators are required 80 to submit a self-declaration form, as outlined in the ICMR guidelines, to the institutional 81 authorities and ethics committee for information. 82

DRAFT

5 | P a g e

Sputum specimens should be collected, as required for routine diagnostic evaluation, from 83 patients who are suspected of having pulmonary TB as per algorithm. Probability of harm or 84 discomfort anticipated in the research is nil or not expected. 85
Enrolment of subjects should be continued till the sample size is met or till the project 86 duration is completed. 87
If additional sputum sample is obtained, written consent must be obtained as per the ICMR 88 National Ethical Guidelines for Biomedical and Health Research Involving Human 89 Participants. The institutional ethics committee of each participating site should be intimated 90 about the study for necessary approval prior to initiating the study. Assent form should be 91 collected in addition to informed consent in case of adolescents (13 to 16 yrs). For children 92 between 7 and 12 years old, oral assent should be obtained in presence of parent or legal 93 guardian. For children under 7 years old, written informed consent should be obtained from 94 parent or legal guardian. 95
The protection of privacy of research participants will be ensured by encrypting the patient 96 identifiers. 97
Patients shall receive the best possible diagnostic work-up as per the routine practice and the 98 National Tuberculosis Elimination Program (NTEP) guidelines. There should not be delay in 99 sending report due to the study.
100
TB treatment decisions should not be made based on the result of the index test under 101 evaluation, but on the basis of the routine clinical and laboratory methods (smear, solid / 102 liquid culture, standard NAAT results, and clinical work-up).
103
Respect for the dignity of research participants should be prioritized. 104
No compensation shall be provided to the participants since there is no additional cost or 105 travel involved in sample collection for the study. Patients should be compensated for travel 106 and time only if they are asked to pay additional visits exclusively for the sake of the study 107 and not during regular treatment visits. 108
Follow-up visits may be required for a very limited number of discrepant patients to exclude 109 TB. 110
Leftover sputum samples and deposits should be stored for resolving discrepancies. One 111 positive culture and two DNA samples per patient should be stored at -80oC for use later. 112

DRAFT

6 | P a g e

All the sites should follow up with all study participants till the final diagnosis is made and 113 the patient should be initiated on appropriate treatment as per NTEP norms. Those found to 114 be M. tuberculosis complex (MTB) positive by standard NAAT test should be started on 115 anti-tuberculosis treatment (ATT) by medical officer of the study site as per NTEP 116 guidelines.
117
The findings of the study should be made accessible through reports. 118

119

120 V. Blinding of Laboratory Staff 121 To ensure rigor of the evaluation process, laboratory staff performing the evaluation should be 122 blinded to the status of the clinical samples. The PI of the evaluation exercise should remain 123 unblinded, i.e., privy to the status of the samples. Another senior laboratory staff selected by the 124 PI may remain unblinded and carry out coding of samples and dispensing them into similar- 125 looking vials to be used for testing, and maintaining the database of results.
126

127 Staff performing the reference test and the test under evaluation (index test), interpretation of the 128 test result, and entering the results against the coded samples in the database, should remain 129 blinded to the status of samples till the completion of evaluation.
130

131 Operators conducting routine laboratory tests (GeneXpert MTB/RIF, MGIT DST, LPA etc.) will 132 not participate in the index test evaluation. Instead, dedicated operators, who are not involved in 133 routine testing and are blinded to the routine test results, will perform the index test. The results 134 will be recorded independently for each test without any patient identifiers. The result sheets will 135 be shared with the investigator for result analysis. The evaluation study data should be analyzed 136 only by the PI of the evaluating lab (Fig. 1). 137

138

139

DRAFT

7 | P a g e

140

141 Figure 1 Blinding in evaluation exercise 142

143 144

DRAFT

8 | P a g e

VI. Procedure 145

146

Preparation of Evaluation sites/laboratories 147  Laboratory must be approved by the NTEP.
148  Accreditation for at least one Quality management system [accreditation for Testing Lab / 149 Calibration Lab (ISO/IES 17025), Medical Lab (ISO 15189), PT provider ISO/IEC 150 17043 or CDSCO approved Reference laboratory]. 151  Three or more sites from different geographical regions should perform clinical 152 validation for representation of population in real world setting. 153

154 2. Study Participants 155 i. For First Line Drug Resistance: People with microbiologically confirmed 156 pulmonary TB by smear and/or NTEP approved NAAT test attending hospital 157 OPDs/Chest clinics/district microscopy centers (DMCs) and Directly Observed 158 Therapy Short Course (DOTS) centers.
159 ii. For Second Line Drug Resistance: All patients with microbiologically confirmed 160 MDR-TB/RR-TB (RIF resistant TB by NAAT test) attending the hospital OPDs/Chest 161 clinics/DMCs/DOTS centers
162 All such consecutive cases (not currently receiving ATT) and willing to provide 163 consent should be enrolled in the study. 164

165 3. Eligibility of Participants 166 Inclusion criteria for testing First Line Drugs 167 i. Individuals positive for TB by smear or any approved NAAT test (Xpert® MTB/RIF) and 168 not receiving ATT
169 ii. Individuals willing to give consent 170 iii. Individuals who are able and willing to give two good quality mucopurulent sputum 171 samples of ≥ 3 ml 172

173

174

DRAFT

9 | P a g e

Inclusion criteria for testing Second Line Drugs 175 i. All patients with microbiologically confirmed MDR-TB/RR-TB (RIF positive by NAAT 176 test) 177 ii. Individuals who are able and willing to give two sputum samples of ≥ 3 ml 178

179 Exclusion criteria 180 i. Individuals on TB treatment for >10 days 181 ii. Individuals not consenting for the study
182 iii. Individuals unable to produce two sputum samples of ≥ 3 ml
183

184 4. Reference and Index tests
185

Index test Reference Test Comparator First Line Drug Resistance New NAAT test for RIF/INH MGIT Culture DST for RIF and INH FL-LPA: GenoType MTBDRplus Second Line Drug Resistance New NAAT test for FQ MGIT Culture DST for Moxifloxacin (0.25, 1 mg) and Levofloxacin (1 mg) SL-LPA: GenoType MTBDRsl

186 5. Sample size
187 Sample size for RIF and INH resistance among TB patients 188 The expected sensitivity of the index test is about 90% with 5 % precision and the expected 189 specificity is 95% with 5% precision. With a confidence interval of 95 % and assuming 10 % 190 loss due to indeterminate results, the sample size required is estimated to be approximately 191 200 patient’s positive each for INH and RIF resistance either alone or in combination. The 192 average prevalence of Isoniazid and Rifampicin are ~18 % and 7.3 % respectively, among the 193 new and previously treated TB patients combined together (Report of drug resistance survey, 194 2014-16). The number needed to screen to obtain 200 drug resistant cases will be 195 approximately 1111 for INH resistance and 2857 for RIF resistance. The participants will be 196 enrolled till the required sample size is achieved for INH and RIF resistance. 197

198

199

DRAFT

10 | P a g e

Sample size for FQ resistance among MDR/RR TB patients 200 The expected sensitivity of the index test for detecting FQ resistance is 90 % with 5 % 201 precision and the expected specificity is 95 % with 5 % precision. Assuming 10 % loss, the 202 sample size required is 200 FQ resistant cases. The prevalence of FQ resistance among 203 MDR/RR TB patients is 20 % (Report of drug resistance survey, 2014-16). Hence, the number 204 needed to screen will be approximately 890. The participants will be enrolled till the required 205 sample size is achieved for FQ resistance. Table 1 shows sample sizes required for RIF, INH 206 and FQ drug resistance. 207

208 Table 1. Sample sizes for RIF, INH and FQ Drug Resistance 209

210

Assumptions for Sensitivity Assumptions for Specificity Sensitivity/Specificity of the new test (%) 90 95 Relative precision (d) (%)
5 5 Desired confidence level (1- alpha) % 95 95 Number of drug resistance (INH and RIF) cases required 178 84 Number of drug resistant cases required with 10 % loss due to indeterminate results
~200 ~93 Number needed to be screened assuming a combined weighted average prevalence of ~18 % for INH resistance among the new and previously treated TB patients 1111 517 Number needed to be screened assuming a combined weighted average prevalence of ~7 % for RIF resistance among the new and previously treated TB patients 2857 1329 Number needed to be screened considering a prevalence of 20 % for FQ resistance among MDR/RR TB 890 465

211 Other disease controls (to check cross-reactivity in real patients) 212 Include people with common alternative diagnoses to mirror programmatic reality and probe 213 false positives. This subset helps characterize clinical exclusivity beyond simple “TB- 214 negative” status: 215

DRAFT

11 | P a g e

216 i. Non-Tuberculous Mycobacteria (Culture or PCR confirmed): ~30 217 ii. Other respiratory diseases [e.g., bacterial pneumonia, chronic obstructive pulmonary 218 disease (COPD), lung cancer, chronic fungal (like Histoplasmosis or Aspergillosis)]: ~30 219 patients combined. 220

221 6. Implementation Plan 222 The samples will be collected and tested as per the routine practice for smear, Xpert 223 MTB/RIF®, LPA, MGIT culture and DST. The samples with positive result for MTB either in 224 smear or NAAT test should be tested for first line drug resistance (RIF and INH). The 225 samples that are positive for MDR/RR (positive for rifampicin resistance by NAAT test) 226 should be used for testing drug resistance for second line drugs. 227

228

229

230 Figure 2. Flowchart for evaluating IVDs for testing drug resistance to RIF and INH 231 among pulmonary TB (PTB) patients
232

DRAFT

12 | P a g e

233

234

235 Figure 3.Flowchart for evaluating IVDs for testing drug resistance to FQ among MDR/RR 236 TB patients
237

238

DRAFT

13 | P a g e

Sample collection, processing and storage 239

240

Two sputum samples each of minimum 3 ml should be collected (one spot and one morning 241 specimen) and sent to laboratory.
242
Approximately 1 ml of sample should be taken from each sample and pooled under sterile 243 conditions (total of 2 ml). 244
Around 1 ml of pooled sample should be tested by the standard NAAT (Xpert MTB/RIF®) 245 and remaining sample used for index test(s).
246
The remaining portion of each sputum sample should be subjected to direct smear and 247 decontamination by NaLC-NaOH method individually.
248
All smear positive or NAAT positive samples will be tested by Line Probe Assay (LPA). 249
The resultant deposit should be used for inoculation into two MGIT960 tubes.
250
All positive cultures should be identified using rapid Immuno-chromatography test (ICT). 251 (Ideally, positive MGIT tubes are tested within 5 days of instrument positivity. Interpretation 252 of the result should be done within 15 minutes). 253
The positive cultures should be tested for drug sensitivity. 254
All sputum samples should be stored at -20oC for later use. Decontaminated sediments and 255 one positive culture per patient should be stored at -80oC, if necessary for later use. 256
Two DNA samples (one DNA sample extracted for index test and one for LPA) per patient 257 should be stored at -20oC till the end of the study for resolution of discrepant results.
258
The index tests should be carried out as per the algorithm (figure 2) and as per the 259 manufacturers’ instructions in the instructions for use (IFU).
260

261 All conventional test procedures for smear, culture (solid and liquid) and Xpert MTB will be 262 performed as per NTEP national laboratory guidelines (CTD, 2016; RNTCP 2009) and laboratory 263 manual of ICMR-NIRT (NIRT, 2010). Standard operating procedures for index test(s) will be 264 provided by the manufacturer(s) including use of positive and negative controls. All procedures 265 for preparation of media, reagents, washing, decontamination, disposal and storage will be 266 performed according to the standard operating procedures (SOP) of ICMR-NIRT (NIRT, 2010) 267 and WHO, (WHO, 2022). 268

269

270

271

DRAFT

14 | P a g e

Laboratory Tests
272

273 i. Smear microscopy: Two direct sputum smear
274 ii. MGIT culture (decontaminated with 1-1.5% final NaOH); Two MGIT tubes (one per 275 specimen) for each patient 276 iii. MGIT drug sensitivity testing (DST) for Rif, INH: Drug sensitivity testing will be carried 277 out from any one positive MGIT culture. 278 iv. MGIT drug sensitivity testing for moxifloxicin (0.25 mg and 1 mg) and levofloxacin (1 279 mg). Drug sensitivity testing should be carried out in from any one positive MGIT culture. 280 v. Speciation of culture: Rapid immunochromatographic test (ICT) of MGIT culture
281 vi. LPA: LPA shall be carried out as per routine practice and as per NTEP guidelines. Direct 282 LPA should be carried out from any one smear positive sample. If the sample is smear 283 negative and culture positive, indirect LPA should be carried out from culture. First line 284 LPA (FL-LPA) will be carried out (Rif and INH resistance) 285 vii. XpertMTB/RIF (one test per patient) 286

287 9. Index test 288 i. Index test will be performed as per manufacturer’s instructions following blinded study 289 protocols. 290 ii. At least 2 different lots of reagents should be tested across the study population to 291 demonstrate consistency of test performance and minimize lot-related bias. 292 iii. The results of the index test will not be disclosed to study participants or clinicians and will 293 not be used to guide treatment decisions. 294

295 10 . Data Analysis and resolution of discrepancy 296 i. If the index test produces error or indeterminate results, then only one repeat is allowed. 297 The results of first test and repeat test should be recorded separately. All 298 Invalids/Indeterminates/errors should be recorded and reported. 299 ii. Results for new patients and previously treated patients should be entered separately. 300 Result analysis will be carried out for these two populations separately as well as 301 combined. 302 iii. A subgroup analysis may be carried out for pediatric population. 303

DRAFT

15 | P a g e

Quality Control (QC) measures 304 All sites should ensure high quality laboratory procedures, data recording and documentation. 305 There should be no deviation from the protocol. All the sites should participate in internal 306 quality control (IQC) and external quality assurance (EQA) for all methods as per the 307 standard manuals of Global Laboratory Initiative (GLI, 2014). 308 Culture: Positive (Reference strain H37Rv or H37Ra) and negative controls for MGIT and 309 LJ cultures would be tested as per NTEP guidelines. MGIT Time to detection QC for MTB 310 reference strain would be performed every month/new lot of reagents/machine service. 311 Sterility and performance testing of culture media would be performed with every new batch 312 or lot. 313 Drug sensitivity testing (DST): Standard ATCC strains should be used for each drug as 314 reference control. QC should be performed whenever a new batch of drugs is prepared, after 315 servicing of the instrument and after long gap of setting up DST. 316 Molecular diagnostics: For molecular diagnostics internal quality control includes control 317 supplied by the manufacturer and control prepared by the lab from the previous testing. The 318 internal control should be used whenever batch of test kit changes, machine is serviced, and 319 newly trained person is introduced into the system.
320

321 VII. Statistical Analysis Plan 322 i. The performance of the diagnostic kits should be evaluated by calculating the sensitivity, 323 specificity, positive predictive value, negative predictive value and accuracy with reference to 324 the gold standard. 95% Confidence interval should be calculated for each of the parameters. 325 ii. The index molecular test will be evaluated for its performance with reference to MGIT DST 326 (for RIF/INH/FQ).
327 iii. Similarly, the performance of NTEP approved molecular test (Xpert MTB/RIF and LPA) 328 should be estimated with reference to MGIT DST. 329 iv. The agreement between the index test and molecular test for drug resistance (LPA) should be 330 calculated using kappa statistic. 331

332

333

DRAFT

16 | P a g e

VIII. Acceptance Criteria 334 Expected minimal sensitivity for MTB and Drug Resistant TB: ≥85 ± 2% 335 Expected minimal specificity for MTB and Drug Resistant TB: ≥95 ± 2% 336 Sample size: ~200 positives for each drug resistance (RIF or INH or FQ etc) (either alone or in 337 combination) and ~ 100 negatives for each drug resistance (RIF or INH or FQ etc). 338

353 References 354

Report of the first national anti-tuberculosis drug resistance survey India, 2014-2016. 355
Technical and operational guidelines for tuberculosis control in India 2016. Central TB 356 Division. 357
RNTCP Standard Operating Procedures for Tuberculosis lab for culture and DST, 2009. 358
Standard Operating Procedures (SOP) for Mycobacteriology laboratory, ICMR-NIRT, 2010. 359
Practical manual on tuberculosis laboratory strengthening, 2022 update. Geneva: World 360 Health Organization; 2022. Licence: CC BY-NC-SA 3.0 IGO. 361
Mycobacteriology laboratory manual, Global laboratory initiative, First edition, April 2014, 362 Stop TB Partnership. 363

364

365

DRAFT

17 | P a g e

PERFORMANCE EVALUATION REPORT FORMAT 366 Performance Evaluation Report For MDR-TB Kit 367

368 Name of the product(Brand/generic)

Name and address of the legal manufacturer

Name and address of the actual manufacturing site

Name and address of the Importer

Name of supplier: Manufacturer/Importer/Port office of CDSCO/State licensing Authority

Lot No /Batch No.:

Product Reference No/Catalogue No

Type of Assay

Kit components

Manufacturing Date

Expiry Date

Pack size (Number of tests per kit)

Intended Use

Number of Tests Received

Regulatory Approval: Import license / Manufacturing license/ Test license

License Number: Issue date: Valid Upto:

Application No.

Sample Panel Sample type

Positive samples (provide details: strong, moderate, weak)

Negative samples (provide detail: clinical/spiked, including cross reactivity panel)

369

370 Results:
371

372 Test Number of samples tested Positive Negative Invalids/Indeterminates/ Error/Contamination (culture) Smear

MGIT culture

Xpert MTB/RIF

Number of samples tested Sensitive Resistant

FL LPA – RIF

FL LPA - INH

DRAFT

18 | P a g e

SL LPA- FQ

MGIT-DST- RIF

MGIT-DST-INH

MGIT-DST-FQ

New IVD- RIF

New IVD-INH

New IVD-FQ

373

Reference assay ……….…………… (MGITDST – RIF/INH/FQ)*

Positive Negative Total Name of MDR-TB kit Positive

Negative

Total

374

Estimate (%) 95% CI Sensitivity

Specificity

375 *Report RIF/INH/FQ as separate tables 376

377 Conclusions: 378

379 o Sensitivity, specificity 380 o Performance: Satisfactory / Not satisfactory 381 (Sensitivity and specificity have been assessed in controlled lab setting using kits provided by the manufacturer from 382 the batch mentioned above using ….. sample. Results should not be extrapolated to other sample types.) 383

384 DISCLAIMERS 385

386

This validation process does not approve / disapprove the kit design 387
This validation process does not certify user friendliness of the kit / assay 388

389 Note: This report is exclusively for ………………………Kit (Lot Nos.……), version …………with the 390 gene targets …………………...manufactured by …………… (Supplied by ……….). 391

392 Evaluation Done on …………………… 393

394 Evaluation Done by …………………………. 395

396

397

398

399 Signature of Director/ Director-In-charge …………………… Seal ………………………… 400

401 ****End of the Report 402

403

ICMR-CDSCO STANDARD PERFORMANCE EVALUATION PROTOCOLS

5 FIELD PERFORMANCE EVALUATION OF IN-VITRO 6 DIAGNOSTICS FOR PULMONARY TUBERCULOSIS 7

ICMR-CDSCO/IVD/TB/PROTOCOLS/2/2025

10 DIVISION OF COMMUNICABLE DISEASES, ICMR IN VITRO DIAGNOSTICS DIVISION, CDSCO AUGUST, 2025 India

DRAFT Field Performance Evaluation of IVD for Pulmonary Tuberculosis

2 | P a g e

12 LIST OF CONTRIBUTORS 13

15 A. Working Group: 16

Dr Gomathi N. Sivaramakrishnan, Former Scientist, ICMR-NIRT, Chennai 17
Dr Sivakumar, Scientist E and Head, Department of Bacteriology, ICMR-NIRT, Chennai 18
Dr Madhumathi J, Scientist D, Division of Communicable Diseases (CD), ICMR, New Delhi 19
Dr Hansraj Choudhary, Scientist C, Division of Communicable Diseases (CD), ICMR, New 20 Delhi 21

23 B. Review Committee: 24

Dr Amita Jain, Professor and Head, Department of Microbiology, KGMU 26
Dr RM Pandey, ICMR- Dr A.S. Paintal Distinguished Scientist Chair 27
Dr. Camilla Rodrigues, Senior Consultant, P.D. Hinduja Hospital, Mumbai 28
Dr Gita Nataraj, Professor Emeritus, Microbiology, Seth GS Medical College and 29 KEM Hospital, Mumbai
30
Dr Ashutosh Aggarwal, Professor and Head, Pulmonary Medicine, PGIMER 31
Dr Venkataraghava Mohan, Professor & Head, Dept. of Community Health and 32 Development, CMC Vellore 33
Mr. Pramod Meshram, Deputy Drugs Controller, Central Drugs Standard Control 34 Organization, Directorate General of Health Services, Ministry of Health and Family 35 36
Dr. Sella Senthil, Assistant Drugs Controller, Central Drugs Standard Control 37 Organization, Directorate General of Health Services, Ministry of Health and Family 38 39
Dr. Nivedita Gupta, Scientist-G and Head of the Division of Communicable Diseases, 40 ICMR Headquarters, Department of Health Research, Ministry of Health and Family 41 42

DRAFT

3 | P a g e

47 Table of Contents 48

7 7 8 8 8 10 11 12 12 12 VII.
Statistical Analysis Plan 13 VIII.
Acceptance Criteria 13 Important Note 14

References 14

Performance Evaluation Report Format 15

DRAFT

4 | P a g e

I. Background 54 CDSCO and ICMR, New Delhi, have aimed at facilitating the availability of Quality-Assured 55 diagnostic kits appropriate for use in India. This protocol gives the methods to be used for 56 evaluating the clinical performance characteristics of nucleic acid amplification based in-vitro 57 diagnostic test in detecting pulmonary tuberculosis. 58

59 II. Purpose 60 To evaluate the clinical performance characteristics of nucleic acid amplification tests (NAAT) 61 for diagnosis of pulmonary Mycobacterium Tuberculosis (MTB) using prospectively collected 62 sputum samples in clinical setting. 63

64 III. Study Design 65 Cross-sectional prospective multi-centric diagnostic accuracy study of IVD for detection of 66 pulmonary TB using Mycobacterium Growth Indicator Tube (MGIT) liquid culture as the 67 microbiological reference standard. 68

69 IV. Ethical Considerations 70

The study should be compliant to the ICMR’s Guidance on Ethical Requirements for 71 Laboratory Validation Testing, 2024. Performance evaluation activities using irreversibly de- 72 identified leftover clinical samples are exempt from ethics approval as per ICMR’s Guidance 73 on Ethical Requirements for Laboratory Validation Testing, 2024. Investigators are required 74 to submit a self-declaration form, as outlined in the ICMR guidelines, to the institutional 75 authorities and ethics committee for information. 76
Sputum specimens should be collected, as required for routine diagnostic evaluation, from 77 patients who are suspected of having pulmonary TB as per algorithm. Probability of harm or 78 discomfort anticipated in the research is nil or not expected. 79
Enrolment of subjects should be continued till the sample size is met or till the project 80 duration is completed. 81
If additional sputum sample is obtained, written consent must be obtained as per the ICMR 82 National Ethical Guidelines for Biomedical and Health Research Involving Human 83 Participants. The institutional ethics committee of each participating site should be intimated 84

DRAFT

5 | P a g e

about the study for necessary approval prior to initiating the study. Assent form should be 85 collected in addition to Informed Consent in case of adolescents (13 to 16 yrs). For children 86 between 7 and 12 years old, oral assent should be obtained in presence of parent or legal 87 guardian. For children under 7 years old, written informed consent should be obtained from 88 parent or legal guardian. 89 5. The protection of privacy of research participants will be ensured by encrypting the patient 90 identifiers. 91 6. Patients shall receive the best possible diagnostic work-up as per the routine practice and the 92 National Tuberculosis Elimination Program (NTEP) guidelines. There should not be delay in 93 sending report due to the study.
94 7. TB treatment decisions should not be made based on the result of the index test under 95 evaluation, but on the basis of the routine clinical and laboratory methods (smear, solid / 96 liquid culture, standard NAAT results, and clinical work-up).
97 8. Respect for the dignity of research participants should be prioritized. 98 9. No compensation shall be provided to the participants since there is no additional cost or 99 travel involved in sample collection for the study. Patients should be compensated for travel 100 and time only if they are asked to pay additional visits exclusively for the sake of the study 101 and not during regular treatment visits. 102 10. Follow-up visits may be required for a very limited number of discrepant patients to exclude 103 TB. 104 11. Leftover sputum samples and deposits should be stored for resolving discrepancies. One 105 positive culture and two DNA samples per patient should be stored at -80oC for use later. 106 12. All the sites should follow up with all study participants till the final diagnosis is made and 107 the patient should be initiated on appropriate treatment as per NTEP norms. Those found to 108 be M. tuberculosis complex (MTB) positive by standard NAAT test should be started on 109 anti-tuberculosis treatment (ATT) by medical officer of the study site as per NTEP 110 guidelines.
111 13. The findings of the study should be made accessible through reports. 112

113

114

115

116

DRAFT

6 | P a g e

V. Blinding of Laboratory Staff 117 To ensure rigor of the evaluation process, laboratory staff performing the evaluation should be 118 blinded to the status of the clinical samples. The PI of the evaluation exercise should remain 119 unblinded, i.e., privy to the status of the samples. Another senior laboratory staff selected by the 120 PI may remain unblinded and carry out coding of samples and dispensing them into similar- 121 looking vials to be used for testing, and maintaining the database of results.
122

123 Staff performing the reference test and the test under evaluation (index test), interpretation of the 124 test result, and entering the results against the coded samples in the database, should remain 125 blinded to the status of samples till the completion of evaluation.
126

127 Operators conducting routine laboratory tests (smear, Xpert MTB/RIF, MGIT culture etc) will 128 not participate in the index test evaluation. Instead, dedicated operators, who are not involved in 129 routine testing and are blinded to the routine test results, will perform the index test. The results 130 will be recorded independently for each test without any patient identifiers. The result sheets will 131 be shared with the investigator for result analysis. The data should be analyzed only by the PI of 132 the evaluating lab (Fig. 1). 133

134

135

DRAFT

7 | P a g e

136

137 Figure 1 Blinding in evaluation exercise 138

139 VI. Procedure 140

141

Preparation of Evaluation sites/laboratories 142  Laboratory must be approved by the National TB Elimination Program (NTEP).
143  Accreditation for at least one Quality management system [accreditation for Testing Lab / 144 Calibration Lab (ISO/IES 17025), Medical Lab (ISO 15189), PT provider ISO/IEC 145 17043 or CDSCO approved Reference laboratory]. 146  Three or more sites from different geographical regions should perform clinical 147 validation for representation of population in real world setting. 148

149 2. Study Participants 150 Individuals with symptoms of presumptive pulmonary TB attending hospital OPDs/Chest 151 clinics/district microscopy centers (DMCs) and Directly Observed Therapy Short Course 152 (DOTS) centers. All such consecutive cases willing to provide consent will be enrolled in the 153 study. 154

DRAFT

8 | P a g e

155

156 3. Eligibility of Participants 157 Inclusion Criteria 158

Individuals positive for TB by smear or any approved NAAT test (Xpert® MTB/RIF) 159
Individuals willing to give consent 160
Individuals who are able and willing to give two good quality mucopurulent sputum 161 samples of ≥ 3 ml 162 Exclusion criteria 163
Individuals on TB treatment for >96 hrs 164
Individuals not consenting for the study
165
Individuals unable to produce two sputum samples of ≥ 3 ml 166

167 4. Reference and Index tests
168 Reference test: Mycobacterium Growth Indicator Tubes (MGIT) liquid culture 169 Comparator: NTEP approved NAAT test (Xpert® MTB/RIF) 170

171 5. Sample size
172 The anticipated sensitivity of an index test is 90 % and with absolute 5 % precision, while the 173 anticipated specificity is 99 per cent with 1 % precision. A higher precision for specificity 174 would be required to minimize false positivity. The minimum sample size requirement has 175 been calculated as ~150 positives and ~470 negatives for MTB by the gold standard culture.
176 With a prevalence of 24 % culture positives among presumptive cases in hospital setting 177 (Penn-Nicholson et al., 2021) and a 5 % loss due to indeterminate results, approximately 610 178 consecutive cases meeting the inclusion and exclusion criteria would be required to be 179 enrolled for the detection of MTB (Jayaprakasam et al., 2024). Enrolment would be continued 180 till the required number of participants is covered. 181 Definition of Presumptive PTB: Patients with any of the following symptoms regardless of duration will be considered to have
‘presumptive TB’: cough for two weeks or more, fever for two weeks or more, night sweats, unintentional weight loss, hemoptysis, chest pain or loss of appetite, with any abnormality in chest radiograph (one or more of the following findings by standardized interpretative criteria: cavitary lesion(s), apical infiltrates, hilar lymphadenopathy, new infiltrates and other suggestive radiological findings).

DRAFT

9 | P a g e

182 The formula for calculating sample size for determining sensitivity/specificity of the index test: 183

184 NSe = [Z (1-α/2)]2 (Se)(1-Se)] 185 d2 186 or 187

188 NSp = [Z (1-α/2)]2 (Sp)(1-Sp)] 189 d2 190 NSe: Sample size for estimating sensitivity,
191 Se: Anticipated sensitivity with reference to culture DST
192 Sp: Anticipated specificity with reference to culture DST
193 Z (1-α/2):1.96 for confidence level of 95%
194 d: Absolute precision 195

196

197

DRAFT

10 | P a g e

Implementation Plan 198

199

200

201 Figure 2 Flowchart for evaluating NAAT test for detection of Mycobacterium 202 Tuberculosis (MTB) among individuals with presumptive pulmonary TB (PTB)
203

204

205

206

DRAFT

11 | P a g e

Sample collection, processing and storage
207

208

Two sputum samples each of minimum 3 ml should be collected (one spot and one morning 209 specimen) and sent to laboratory.
210
Approximately 1 ml of sample should be taken from each sample and pooled under sterile 211 conditions (total of 2 ml). 212
Around 1 ml of pooled sample should be tested by the standard NAAT (Xpert MTB/RIF®) 213 and remaining sample used for index test(s).
214
The remaining portion of each sputum sample should be subjected to direct smear and 215 decontamination by NaLC-NaOH method individually.
216
The resultant deposit should be used for inoculation into two MGIT960 tubes.
217
All positive cultures should be identified using rapid Immuno-chromatography test (ICT). 218 (Ideally, positive MGIT tubes are tested within 5 days of instrument positivity. Interpretation 219 of the result should be done within 15 minutes). 220
All sputum samples should be stored at -20oC for later use. Decontaminated sediments and 221 one positive culture per patient should be stored at -80oC, if necessary for later use. 222
Two DNA samples per patient should be stored at -20oC till the end of the study for 223 resolution of discrepant results.
224
The index tests should be carried out as per the algorithm (figure 2) and as per the 225 manufacturers’ instructions in the instructions for use (IFU).
226

227 All conventional test procedures for smear, culture (solid and liquid) and Xpert MTB will be 228 performed as per NTEP national laboratory guidelines (CTD, 2016; RNTCP 2009) and laboratory 229 manual of ICMR-NIRT (NIRT, 2010). Standard operating procedures for index test(s) will be 230 provided by the manufacturer(s) including use of positive and negative controls. All procedures 231 for preparation of media, reagents, washing, decontamination, disposal and storage will be 232 performed according to the standard operating procedures (SOP) of ICMR-NIRT (NIRT, 2010) 233 and WHO, (WHO, 2022). 234

235

236

237

238

239

240

DRAFT

12 | P a g e

Laboratory Tests
241

242 i. Smear microscopy: Two direct sputum smear
243 ii. MGIT culture (decontaminated with 1-1.5% final NaOH); Two MGIT tubes (one per 244 specimen) for each patient 245 iii. Speciation of culture: Rapid immunochromatographic test (ICT) of MGIT culture
246 iv. Xpert MTB/RIF (one test per patient) 247

248 9. Data Analysis and resolution of discrepancy 249 i. If the index test produces error or indeterminate results, then only one repeat is allowed. 250 The results of first test and repeat test should be recorded separately.
251 ii. All Invalids/Indeterminates/errors should be recorded and reported. 252 iii. A subgroup analysis may be carried out for pediatric population. 253

254 10. Quality Control (QC) measures 255 All sites should ensure high quality of laboratory procedures, data recording and 256 documentation. There should be no deviation from the protocol. All the sites should 257 participate in internal quality control (IQC) and external quality assurance (EQA) for all 258 methods as per the standard manuals of Global Laboratory Initiative (GLI, 2014). 259 Culture: Positive (Reference strain H37Rv or H37Ra) and negative controls for MGIT and 260 LJ cultures would be tested as per NTEP guidelines. MGIT Time to detection QC for MTB 261 reference strain would be performed every month/new lot of reagents/machine service. 262 Sterility and performance testing of culture media would be performed with every new batch 263 or lot. 264 Smear: Smear QC should be performed as per NTEP guidelines at regular intervals and with 265 new lot of reagents.
266 ICT Identification of MTB complex: Culture of M. tuberculosis reference strain in MGIT 267 broth should be used as positive control. Culture of Mycobacteria other than tuberculosis 268 (e.g., a well characterized strain of M. avium complex/M.kansasii) in MGIT broth should be 269 used as negative control. QC for ICT should be performed every 3 months. 270 Molecular diagnostics: For molecular diagnostics internal quality control includes control 271 supplied by the manufacturer and control prepared by the lab from the previous testing. The 272

DRAFT

13 | P a g e

internal control should be used whenever batch of test kit changes, machine is serviced, and 273 newly trained person is introduced into the system.
274 Avoiding Cross-contamination: Unidirectional workflow: The workflow of a molecular lab 275 should be in one direction only. PCR master mix reagents and samples that may contain 276 templates for PCR should be prepared in the pre-PCR room only. Tubes that have undergone 277 amplification in the post-PCR room contain amplicons and will not be opened or introduced 278 in the pre-PCR room. Consumables and PPE (lab coats, gloves, goggles, etc.) that have been 279 used in the post-PCR room should not be placed back in the pre-PCR room without thorough 280 decontamination. Aerosol resistant pipettes will be used for all procedures and standard 281 aseptic cleaning technique should be carried out before and after PCR for work surface, bench 282 top and equipment. 283

284 VII. Statistical Analysis Plan 285

286 i. The performance of the diagnostic kits should be evaluated by calculating the sensitivity, 287 specificity, positive predictive value, negative predictive value and accuracy with reference 288 to the gold standard. 95% Confidence interval should be calculated for each of the 289 parameters. 290 ii. The index molecular test should be evaluated for its performance with reference to the 291 MGIT culture.
292 iii. Similarly, the performance of standard molecular test (Xpert MTB/RIF) should be estimated 293 with reference to MGIT culture. 294 iv. The sensitivity and specificity of index test vs MGIT culture should be compared with that 295 of Xpert® MTB/RIF Vs MGIT culture. 296 v. The agreement between the index test and standard NAAT test (Xpert MTB/RIF) should be 297 calculated with kappa statistic. 298

299 VIII. Acceptance Criteria 300 Expected sensitivity: ≥85 ± 2% 301 Expected specificity: ≥95 ± 2% 302 Sample size: ~150 MTB positives and ~470 MTB negatives by MGIT culture 303

304

DRAFT

14 | P a g e

IMPORTANT NOTE 305 After following due procedure as defined in this document, once any kit is found to be Not 306 of Standard Quality, thereafter, no request for repeat testing of the same kit will be 307 acceptable.
308

309 Any request of re-validation from the same manufacturer for the same test type will only 310 be entertained after a minimum of 3 months and only if a high-level technical summary of 311 modifications or functional improvements to the kit design is submitted, without explicit 312 disclosure of proprietary information. 313

314 Clinical samples are precious, therefore, repeat evaluation of a kit using the same/ different 315 well-characterized sample panel at a different laboratory may be considered only for kits 316 which claim high performance characteristics (sensitivity and specificity 95% and above), 317 but which fail the performance evaluation by a margin of 5%. 318

319 References 320

Penn-Nicholson, A., Gomathi, S. N., Ugarte-Gil, C., Meaza, A., Lavu, E., Patel, P., 321 Choudhury, B., Rodrigues, C., Chadha, S., Kazi, M., Macé, A., Nabeta, P., Boehme, C., 322 Gangakhedkar, R. R., Sarin, S., Tesfaye, E., Gotuzzo, E., du Cros, P., Tripathy, S., Ruhwald, 323 M., … Members of the Truenat Trial Consortium: (2021). A prospective multicentre 324 diagnostic accuracy study for the Truenat tuberculosis assays. The European respiratory 325 journal, 58(5), 2100526.
326
Jayaprakasam, M., Pandey, R. M., Choudhary, H., Shanmugam, S., Sivaramakrishnan, G. 327 N., & Gupta, N. (2024). Evaluation of molecular diagnostic test for detection of adult 328 pulmonary tuberculosis: A generic protocol. The Indian journal of medical research, 159(2), 329 246–253.
330
Technical and operational guidelines for tuberculosis control in India 2016. Central TB 331 Division. 332
RNTCP Standard Operating Procedures for Tuberculosis lab for culture and DST, 2009. 333
Standard Operating Procedures (SOP) for Mycobacteriology laboratory, ICMR-NIRT, 2010. 334
Practical manual on tuberculosis laboratory strengthening, 2022 update. Geneva: World 335 Health Organization; 2022. Licence: CC BY-NC-SA 3.0 IGO. 336
Mycobacteriology laboratory manual, Global laboratory initiative, First edition, April 2014, 337 Stop TB Partnership. 338 339

DRAFT

15 | P a g e

PERFORMANCE EVALUATION REPORT FORMAT 340

341 Performance Evaluation Report For MTB Kit 342

343 Name of the product (Brand/generic)

Name and address of the legal manufacturer

Name and address of the actual manufacturing site

Name and address of the Importer

Name of supplier: Manufacturer/Importer/Port office of CDSCO/State licensing Authority

Lot No /Batch No.:

Product Reference No/Catalogue No

Type of Assay

Kit components

Manufacturing Date

Expiry Date

Pack size (Number of tests per kit)

Intended Use

Number of Tests Received

Regulatory Approval: Import license / Manufacturing license/ Test license

License Number: Issue date: Valid Upto:

Application No.

Sample Panel Sample type

Positive samples (provide details: strong, moderate, weak)

Negative samples (provide detail: clinical/spiked, including cross reactivity panel)

344 Results:
345

346 Test Number of samples tested Positive Negative Invalids/ Indeterminates/Error/ Contamination (culture) Smear

MGIT culture

Xpert MTB/RIF

New MTB kit

347

348

DRAFT

16 | P a g e

Reference assay ……….…………… (MGIT culture)

Positive Negative Total Name of MTB kit Positive

Negative

Total

349

Estimate (%) 95% CI Sensitivity

Specificity

350

351

352 Conclusions: 353

354 o Sensitivity, specificity 355 o Performance: Satisfactory / Not satisfactory 356 (Sensitivity and specificity have been assessed in controlled lab setting using kits provided by the manufacturer from 357 the batch mentioned above using ….. sample. Results should not be extrapolated to other sample types.) 358

359 DISCLAIMERS 360

361

This validation process does not approve / disapprove the kit design 362
This validation process does not certify user friendliness of the kit / assay 363

364 Note: This report is exclusively for ………………………Kit (Lot No……), version …………with the 365 gene targets …………………...manufactured by …………… (Supplied by ……….). 366

367 Evaluation Done on …………………… 368

369 Evaluation Done by …………………………. 370

371

372

373

374 Signature of Director/ Director-In-charge …………………… Seal ………………………… 375

376 ****End of the Report 377

378

379

380

381

Verbatim extracted text (OCR/PDF). Older scans and tables may show extraction artifacts — verify against the original for anything you act on.

Analysis

No analysis has been generated for this document yet.