Propensity score-based approaches to confounding by indication in individual patient data meta-analysis: Non-standardized treatment for multidrug resistant tuberculosis

Gregory J. Fox, Andrea Benedetti, Carole D. Mitnick, Madhukar Pai, Dick Menzies, S. Ahuja, D. Ashkin, M. Avendaño, R. Banerjee, M. Bauer, M. Becerra, M. Burgos, R. Centis, E. D. Chan, C. Y. Chiang, F. Cobelens, H. Cox, L. D'Ambrosio, W. C M De Lange, K. DeRiemerD. Enarson, D. Falzon, K. Flanagan, J. Flood, N. Gandhi, L. Garcia-Garcia, R. M. Granich, M. G. Hollm-Delgado, T. H. Holtz, P. Hopewell, M. Iseman, L. G. Jarlsberg, S. Keshavjee, H. R. Kim, J. Lancaster, C. Lange, V. Leimane, C. C. Leung, W. J. Koh, J. Li, D. Menzies, G. B. Migliori, M. Narita, E. Nathanson, R. Odendaal, P. O'Riordan, M. Pai, D. Palmero, S. K. Park, G. Pasvol, J. Pena, C. Pérez-Guzmán, A. Ponce-de-Leon, M. I D Quelapio, H. T. Quy, V. Riekstina, J. Robert, S. Royce, M. Salim, H. S. Schaaf, K. J. Seung, L. Shah, K. Shean, T. S. Shim, S. S. Shin, Y. Shiraishi, J. Sifuentes-Osornio, G. Sotgiu, M. J. Strand, S. W. Sung, P. Tabarsi, T. E. Tupasi, M. H. Vargas, R. Van Altena, P. Viiklepp, M. Van Der Walt, T. S. Van Der Werf, J. Westenhouse, W. W. Yew, J. J. Yim

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Background: In the absence of randomized clinical trials, meta-analysis of individual patient data (IPD) from observational studies may provide the most accurate effect estimates for an intervention. However, confounding by indication remains an important concern that can be addressed by incorporating individual patient covariates in different ways. We compared different analytic approaches to account for confounding in IPD from patients treated for multidrug resistant tuberculosis (MDR-TB). Methods: Two antibiotic classes were evaluated, fluoroquinolones-considered the cornerstone of effective MDR-TB treatment-and macrolides, which are known to be safe, yet are ineffective in vitro. The primary outcome was treatment success against treatment failure, relapse or death. Effect estimates were obtained using multivariable and propensity-score based approaches. Results: Fluoroquinolone antibiotics were used in 28 included studies, within which 6,612 patients received a fluoroquinolone and 723 patients did not. Macrolides were used in 15 included studies, within which 459 patients received this class of antibiotics and 3,670 did not. Both standard multivariable regression and propensity score-based methods resulted in similar effect estimates for early and late generation fluoroquinolones, while macrolide antibiotics use was associated with reduced treatment success. Conclusions: In this individual patient data meta-analysis, standard multivariable and propensity-score based methods of adjusting for individual patient covariates for observational studies yielded produced similar effect estimates. Even when adjustment is made for potential confounding, interpretation of adjusted estimates must still consider the potential for residual bias.

Original languageEnglish (US)
Article numbere0151724
JournalPLoS One
Volume11
Issue number3
DOIs
StatePublished - Mar 1 2016
Externally publishedYes

Fingerprint

Multidrug-Resistant Tuberculosis
Propensity Score
Fluoroquinolones
Metadata
tuberculosis
meta-analysis
Meta-Analysis
Macrolides
Anti-Bacterial Agents
fluoroquinolones
macrolides
Therapeutics
antibiotics
observational studies
Observational Studies
macrolide antibiotics
randomized clinical trials
relapse
Treatment Failure
Randomized Controlled Trials

ASJC Scopus subject areas

  • Medicine(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Propensity score-based approaches to confounding by indication in individual patient data meta-analysis : Non-standardized treatment for multidrug resistant tuberculosis. / Fox, Gregory J.; Benedetti, Andrea; Mitnick, Carole D.; Pai, Madhukar; Menzies, Dick; Ahuja, S.; Ashkin, D.; Avendaño, M.; Banerjee, R.; Bauer, M.; Becerra, M.; Burgos, M.; Centis, R.; Chan, E. D.; Chiang, C. Y.; Cobelens, F.; Cox, H.; D'Ambrosio, L.; De Lange, W. C M; DeRiemer, K.; Enarson, D.; Falzon, D.; Flanagan, K.; Flood, J.; Gandhi, N.; Garcia-Garcia, L.; Granich, R. M.; Hollm-Delgado, M. G.; Holtz, T. H.; Hopewell, P.; Iseman, M.; Jarlsberg, L. G.; Keshavjee, S.; Kim, H. R.; Lancaster, J.; Lange, C.; Leimane, V.; Leung, C. C.; Koh, W. J.; Li, J.; Menzies, D.; Migliori, G. B.; Narita, M.; Nathanson, E.; Odendaal, R.; O'Riordan, P.; Pai, M.; Palmero, D.; Park, S. K.; Pasvol, G.; Pena, J.; Pérez-Guzmán, C.; Ponce-de-Leon, A.; Quelapio, M. I D; Quy, H. T.; Riekstina, V.; Robert, J.; Royce, S.; Salim, M.; Schaaf, H. S.; Seung, K. J.; Shah, L.; Shean, K.; Shim, T. S.; Shin, S. S.; Shiraishi, Y.; Sifuentes-Osornio, J.; Sotgiu, G.; Strand, M. J.; Sung, S. W.; Tabarsi, P.; Tupasi, T. E.; Vargas, M. H.; Van Altena, R.; Viiklepp, P.; Van Der Walt, M.; Van Der Werf, T. S.; Westenhouse, J.; Yew, W. W.; Yim, J. J.

In: PLoS One, Vol. 11, No. 3, e0151724, 01.03.2016.

Research output: Contribution to journalArticle

Fox, GJ, Benedetti, A, Mitnick, CD, Pai, M, Menzies, D, Ahuja, S, Ashkin, D, Avendaño, M, Banerjee, R, Bauer, M, Becerra, M, Burgos, M, Centis, R, Chan, ED, Chiang, CY, Cobelens, F, Cox, H, D'Ambrosio, L, De Lange, WCM, DeRiemer, K, Enarson, D, Falzon, D, Flanagan, K, Flood, J, Gandhi, N, Garcia-Garcia, L, Granich, RM, Hollm-Delgado, MG, Holtz, TH, Hopewell, P, Iseman, M, Jarlsberg, LG, Keshavjee, S, Kim, HR, Lancaster, J, Lange, C, Leimane, V, Leung, CC, Koh, WJ, Li, J, Menzies, D, Migliori, GB, Narita, M, Nathanson, E, Odendaal, R, O'Riordan, P, Pai, M, Palmero, D, Park, SK, Pasvol, G, Pena, J, Pérez-Guzmán, C, Ponce-de-Leon, A, Quelapio, MID, Quy, HT, Riekstina, V, Robert, J, Royce, S, Salim, M, Schaaf, HS, Seung, KJ, Shah, L, Shean, K, Shim, TS, Shin, SS, Shiraishi, Y, Sifuentes-Osornio, J, Sotgiu, G, Strand, MJ, Sung, SW, Tabarsi, P, Tupasi, TE, Vargas, MH, Van Altena, R, Viiklepp, P, Van Der Walt, M, Van Der Werf, TS, Westenhouse, J, Yew, WW & Yim, JJ 2016, 'Propensity score-based approaches to confounding by indication in individual patient data meta-analysis: Non-standardized treatment for multidrug resistant tuberculosis', PLoS One, vol. 11, no. 3, e0151724. https://doi.org/10.1371/journal.pone.0151724
Fox, Gregory J. ; Benedetti, Andrea ; Mitnick, Carole D. ; Pai, Madhukar ; Menzies, Dick ; Ahuja, S. ; Ashkin, D. ; Avendaño, M. ; Banerjee, R. ; Bauer, M. ; Becerra, M. ; Burgos, M. ; Centis, R. ; Chan, E. D. ; Chiang, C. Y. ; Cobelens, F. ; Cox, H. ; D'Ambrosio, L. ; De Lange, W. C M ; DeRiemer, K. ; Enarson, D. ; Falzon, D. ; Flanagan, K. ; Flood, J. ; Gandhi, N. ; Garcia-Garcia, L. ; Granich, R. M. ; Hollm-Delgado, M. G. ; Holtz, T. H. ; Hopewell, P. ; Iseman, M. ; Jarlsberg, L. G. ; Keshavjee, S. ; Kim, H. R. ; Lancaster, J. ; Lange, C. ; Leimane, V. ; Leung, C. C. ; Koh, W. J. ; Li, J. ; Menzies, D. ; Migliori, G. B. ; Narita, M. ; Nathanson, E. ; Odendaal, R. ; O'Riordan, P. ; Pai, M. ; Palmero, D. ; Park, S. K. ; Pasvol, G. ; Pena, J. ; Pérez-Guzmán, C. ; Ponce-de-Leon, A. ; Quelapio, M. I D ; Quy, H. T. ; Riekstina, V. ; Robert, J. ; Royce, S. ; Salim, M. ; Schaaf, H. S. ; Seung, K. J. ; Shah, L. ; Shean, K. ; Shim, T. S. ; Shin, S. S. ; Shiraishi, Y. ; Sifuentes-Osornio, J. ; Sotgiu, G. ; Strand, M. J. ; Sung, S. W. ; Tabarsi, P. ; Tupasi, T. E. ; Vargas, M. H. ; Van Altena, R. ; Viiklepp, P. ; Van Der Walt, M. ; Van Der Werf, T. S. ; Westenhouse, J. ; Yew, W. W. ; Yim, J. J. / Propensity score-based approaches to confounding by indication in individual patient data meta-analysis : Non-standardized treatment for multidrug resistant tuberculosis. In: PLoS One. 2016 ; Vol. 11, No. 3.
@article{ae986731fa574a9fac1985fd2c98fecf,
title = "Propensity score-based approaches to confounding by indication in individual patient data meta-analysis: Non-standardized treatment for multidrug resistant tuberculosis",
abstract = "Background: In the absence of randomized clinical trials, meta-analysis of individual patient data (IPD) from observational studies may provide the most accurate effect estimates for an intervention. However, confounding by indication remains an important concern that can be addressed by incorporating individual patient covariates in different ways. We compared different analytic approaches to account for confounding in IPD from patients treated for multidrug resistant tuberculosis (MDR-TB). Methods: Two antibiotic classes were evaluated, fluoroquinolones-considered the cornerstone of effective MDR-TB treatment-and macrolides, which are known to be safe, yet are ineffective in vitro. The primary outcome was treatment success against treatment failure, relapse or death. Effect estimates were obtained using multivariable and propensity-score based approaches. Results: Fluoroquinolone antibiotics were used in 28 included studies, within which 6,612 patients received a fluoroquinolone and 723 patients did not. Macrolides were used in 15 included studies, within which 459 patients received this class of antibiotics and 3,670 did not. Both standard multivariable regression and propensity score-based methods resulted in similar effect estimates for early and late generation fluoroquinolones, while macrolide antibiotics use was associated with reduced treatment success. Conclusions: In this individual patient data meta-analysis, standard multivariable and propensity-score based methods of adjusting for individual patient covariates for observational studies yielded produced similar effect estimates. Even when adjustment is made for potential confounding, interpretation of adjusted estimates must still consider the potential for residual bias.",
author = "Fox, {Gregory J.} and Andrea Benedetti and Mitnick, {Carole D.} and Madhukar Pai and Dick Menzies and S. Ahuja and D. Ashkin and M. Avenda{\~n}o and R. Banerjee and M. Bauer and M. Becerra and M. Burgos and R. Centis and Chan, {E. D.} and Chiang, {C. Y.} and F. Cobelens and H. Cox and L. D'Ambrosio and {De Lange}, {W. C M} and K. DeRiemer and D. Enarson and D. Falzon and K. Flanagan and J. Flood and N. Gandhi and L. Garcia-Garcia and Granich, {R. M.} and Hollm-Delgado, {M. G.} and Holtz, {T. H.} and P. Hopewell and M. Iseman and Jarlsberg, {L. G.} and S. Keshavjee and Kim, {H. R.} and J. Lancaster and C. Lange and V. Leimane and Leung, {C. C.} and Koh, {W. J.} and J. Li and D. Menzies and Migliori, {G. B.} and M. Narita and E. Nathanson and R. Odendaal and P. O'Riordan and M. Pai and D. Palmero and Park, {S. K.} and G. Pasvol and J. Pena and C. P{\'e}rez-Guzm{\'a}n and A. Ponce-de-Leon and Quelapio, {M. I D} and Quy, {H. T.} and V. Riekstina and J. Robert and S. Royce and M. Salim and Schaaf, {H. S.} and Seung, {K. J.} and L. Shah and K. Shean and Shim, {T. S.} and Shin, {S. S.} and Y. Shiraishi and J. Sifuentes-Osornio and G. Sotgiu and Strand, {M. J.} and Sung, {S. W.} and P. Tabarsi and Tupasi, {T. E.} and Vargas, {M. H.} and {Van Altena}, R. and P. Viiklepp and {Van Der Walt}, M. and {Van Der Werf}, {T. S.} and J. Westenhouse and Yew, {W. W.} and Yim, {J. J.}",
year = "2016",
month = "3",
day = "1",
doi = "10.1371/journal.pone.0151724",
language = "English (US)",
volume = "11",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "3",

}

TY - JOUR

T1 - Propensity score-based approaches to confounding by indication in individual patient data meta-analysis

T2 - Non-standardized treatment for multidrug resistant tuberculosis

AU - Fox, Gregory J.

AU - Benedetti, Andrea

AU - Mitnick, Carole D.

AU - Pai, Madhukar

AU - Menzies, Dick

AU - Ahuja, S.

AU - Ashkin, D.

AU - Avendaño, M.

AU - Banerjee, R.

AU - Bauer, M.

AU - Becerra, M.

AU - Burgos, M.

AU - Centis, R.

AU - Chan, E. D.

AU - Chiang, C. Y.

AU - Cobelens, F.

AU - Cox, H.

AU - D'Ambrosio, L.

AU - De Lange, W. C M

AU - DeRiemer, K.

AU - Enarson, D.

AU - Falzon, D.

AU - Flanagan, K.

AU - Flood, J.

AU - Gandhi, N.

AU - Garcia-Garcia, L.

AU - Granich, R. M.

AU - Hollm-Delgado, M. G.

AU - Holtz, T. H.

AU - Hopewell, P.

AU - Iseman, M.

AU - Jarlsberg, L. G.

AU - Keshavjee, S.

AU - Kim, H. R.

AU - Lancaster, J.

AU - Lange, C.

AU - Leimane, V.

AU - Leung, C. C.

AU - Koh, W. J.

AU - Li, J.

AU - Menzies, D.

AU - Migliori, G. B.

AU - Narita, M.

AU - Nathanson, E.

AU - Odendaal, R.

AU - O'Riordan, P.

AU - Pai, M.

AU - Palmero, D.

AU - Park, S. K.

AU - Pasvol, G.

AU - Pena, J.

AU - Pérez-Guzmán, C.

AU - Ponce-de-Leon, A.

AU - Quelapio, M. I D

AU - Quy, H. T.

AU - Riekstina, V.

AU - Robert, J.

AU - Royce, S.

AU - Salim, M.

AU - Schaaf, H. S.

AU - Seung, K. J.

AU - Shah, L.

AU - Shean, K.

AU - Shim, T. S.

AU - Shin, S. S.

AU - Shiraishi, Y.

AU - Sifuentes-Osornio, J.

AU - Sotgiu, G.

AU - Strand, M. J.

AU - Sung, S. W.

AU - Tabarsi, P.

AU - Tupasi, T. E.

AU - Vargas, M. H.

AU - Van Altena, R.

AU - Viiklepp, P.

AU - Van Der Walt, M.

AU - Van Der Werf, T. S.

AU - Westenhouse, J.

AU - Yew, W. W.

AU - Yim, J. J.

PY - 2016/3/1

Y1 - 2016/3/1

N2 - Background: In the absence of randomized clinical trials, meta-analysis of individual patient data (IPD) from observational studies may provide the most accurate effect estimates for an intervention. However, confounding by indication remains an important concern that can be addressed by incorporating individual patient covariates in different ways. We compared different analytic approaches to account for confounding in IPD from patients treated for multidrug resistant tuberculosis (MDR-TB). Methods: Two antibiotic classes were evaluated, fluoroquinolones-considered the cornerstone of effective MDR-TB treatment-and macrolides, which are known to be safe, yet are ineffective in vitro. The primary outcome was treatment success against treatment failure, relapse or death. Effect estimates were obtained using multivariable and propensity-score based approaches. Results: Fluoroquinolone antibiotics were used in 28 included studies, within which 6,612 patients received a fluoroquinolone and 723 patients did not. Macrolides were used in 15 included studies, within which 459 patients received this class of antibiotics and 3,670 did not. Both standard multivariable regression and propensity score-based methods resulted in similar effect estimates for early and late generation fluoroquinolones, while macrolide antibiotics use was associated with reduced treatment success. Conclusions: In this individual patient data meta-analysis, standard multivariable and propensity-score based methods of adjusting for individual patient covariates for observational studies yielded produced similar effect estimates. Even when adjustment is made for potential confounding, interpretation of adjusted estimates must still consider the potential for residual bias.

AB - Background: In the absence of randomized clinical trials, meta-analysis of individual patient data (IPD) from observational studies may provide the most accurate effect estimates for an intervention. However, confounding by indication remains an important concern that can be addressed by incorporating individual patient covariates in different ways. We compared different analytic approaches to account for confounding in IPD from patients treated for multidrug resistant tuberculosis (MDR-TB). Methods: Two antibiotic classes were evaluated, fluoroquinolones-considered the cornerstone of effective MDR-TB treatment-and macrolides, which are known to be safe, yet are ineffective in vitro. The primary outcome was treatment success against treatment failure, relapse or death. Effect estimates were obtained using multivariable and propensity-score based approaches. Results: Fluoroquinolone antibiotics were used in 28 included studies, within which 6,612 patients received a fluoroquinolone and 723 patients did not. Macrolides were used in 15 included studies, within which 459 patients received this class of antibiotics and 3,670 did not. Both standard multivariable regression and propensity score-based methods resulted in similar effect estimates for early and late generation fluoroquinolones, while macrolide antibiotics use was associated with reduced treatment success. Conclusions: In this individual patient data meta-analysis, standard multivariable and propensity-score based methods of adjusting for individual patient covariates for observational studies yielded produced similar effect estimates. Even when adjustment is made for potential confounding, interpretation of adjusted estimates must still consider the potential for residual bias.

UR - http://www.scopus.com/inward/record.url?scp=84977489561&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84977489561&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0151724

DO - 10.1371/journal.pone.0151724

M3 - Article

C2 - 27022741

AN - SCOPUS:84977489561

VL - 11

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 3

M1 - e0151724

ER -