| Title: | Network Meta-Analysis Combining Survival and Count Outcomes |
|---|---|
| Description: | Network meta-analysis for survival outcome data often involves several studies only involve dichotomized outcomes (e.g., the numbers of event and sample sizes of individual arms). To combine these different outcome data, Woods et al. (2010) <doi:10.1186/1471-2288-10-54> proposed a Bayesian approach using complicated hierarchical models. Besides, frequentist approaches have been alternative standard methods for the statistical analyses of network meta-analysis, and the methodology has been well established. We proposed an easy-to-implement method for the network meta-analysis based on the frequentist framework in Noma and Maruo (2025) <doi:10.1101/2025.01.23.25321051>. This package involves some convenient functions to implement the simple synthesis method. |
| Authors: | Hisashi Noma [aut, cre]
|
| Maintainer: | Hisashi Noma <[email protected]> |
| License: | GPL-3 |
| Version: | 1.1-1 |
| Built: | 2025-02-01 05:32:03 UTC |
| Source: | https://github.com/cran/survNMA |
Network meta-analysis for survival outcome data often involves several studies only involve dichotomized outcomes (e.g., the numbers of event and sample sizes of individual arms). To combine these different outcome data, Woods et al. (2010) proposed a Bayesian approach using complicated hierarchical models. Besides, frequentist approaches have been alternative standard methods for the statistical analyses of network meta-analysis, and the methodology has been well established. We proposed an easy-to-implement method for the network meta-analysis based on the frequentist framework in Noma and Maruo (2025). This package involves some convenient functions to implement the simple synthesis method.
Noma, H. and Maruo, K. (2025). Network meta-analysis combining survival and count outcome data: A simple frequentist approach. medRxiv: 10.1101/2025.01.23.25321051.
Log hazard ratio estimate and its standard error estimate are calculated from dichotomized data.
calcHR(d1,n1,d0,n0)calcHR(d1,n1,d0,n0)
d1 |
The number of events in arm 1. |
n1 |
The number of sample size in arm 1. |
d0 |
The number of events in arm 2. |
n0 |
The number of sample size in arm 2. |
TE: The log hazard ratio estimate.
seTE: Standard error estimate for the log hazard ratio.
Noma, H. and Maruo, K. (2025). Network meta-analysis combining survival and count outcome data: A simple frequentist approach. medRxiv: 10.1101/2025.01.23.25321051.
Salika, T., Turner, R. M., Fisher, D., Tierney, J. F. and White, I. R. (2022). Implications of analysing time-to-event outcomes as binary in meta-analysis: empirical evidence from the Cochrane Database of Systematic Reviews. BMC Medical Research Methodology 22, 73.
calcHR(1,229,1,227) calcHR(4,374,7,361) calcHR(3,372,7,361) calcHR(2,358,7,361)calcHR(1,229,1,227) calcHR(4,374,7,361) calcHR(3,372,7,361) calcHR(2,358,7,361)
In network meta-analysis, standard error estimates cannot often be obtained for some contrasts of arms in multi-arm trials. This function calculate the standard error estimate of a contrast measure (e.g., log hazard ratio) estimator from partially obtained summary data in multi-arm trials. Without loss of generality, we consider three arms 0, 1 and 2 in the corresponding trial, and suppose the standard error of the contrast measure comparing the arms 1 vs. 2 is not obtained; however, those comparing the arms 1 vs. 0 and 2 vs. 0 are obtained. We can calculate the standard error estimate comparing the arms 1 vs. 2 from the partially available data.
calcse(se1,se2,n1,n2,n0)calcse(se1,se2,n1,n2,n0)
se1 |
The standard error estimate of a contrast measure (e.g., log hazard ratio) estimator for arms 1 vs. 0. |
se2 |
The standard error estimate of a contrast measure (e.g., log hazard ratio) estimator for arms 2 vs. 0. |
n1 |
The sample size of arm 1. |
n2 |
The sample size of arm 2. |
n0 |
The sample size of arm 0. |
The standard error estimate of a contrast measure (e.g., log hazard ratio) estimator for arms 1 vs. 2.
Noma, H. and Maruo, K. (2025). Network meta-analysis combining survival and count outcome data: A simple frequentist approach. medRxiv: 10.1101/2025.01.23.25321051.
Woods, B. S., Hawkins, N., Scott, D. A. (2010). Network meta-analysis on the log-hazard scale, combining count and hazard ratio statistics accounting for multi-arm trials: a tutorial. BMC Medical Research Methodology 10, 54.
calcse(0.096,0.092,1521,1534,1524)calcse(0.096,0.092,1521,1534,1524)
Merging two dataset for the analysis by netmeta package (e.g., an output object by pairwiseHR for dichotomized outcome dataset and the survival outcome dataset). The output object can be straightforwardly applied to the netmeta function.
combine2(data1, data2)combine2(data1, data2)
data1 |
Dataset 1. |
data2 |
Dataset 2. |
studlab: ID variable of studies.
treat1: The treatment of arm 1.
treat2: The treatment of arm 2.
TE: The effect measure estimate.
seTE: Standard error estimate for the effect measure estimator.
Noma, H. and Maruo, K. (2025). Network meta-analysis combining survival and count outcome data: A simple frequentist approach. medRxiv: 10.1101/2025.01.23.25321051.
data(woods1) data(woods2) woods3 <- pairwiseHR(treat, studlab=study, event=d, n, data=woods2) combine2(woods1, woods3)data(woods1) data(woods2) woods3 <- pairwiseHR(treat, studlab=study, event=d, n, data=woods2) combine2(woods1, woods3)
Log hazard ratio estimate and its standard error estimate are calculated from dichotomized dataset simultaneously.
pairwiseHR(treat, studlab, event, n, data)pairwiseHR(treat, studlab, event, n, data)
treat |
The treatments of individual arms. |
studlab |
ID variable of studies. |
event |
The number of events of individual arms. |
n |
The number of sample size of individual arms. |
data |
The dataset object. |
studlab: ID variable of studies.
treat1: The treatment of arm 1.
treat2: The treatment of arm 2.
TE: The log hazard ratio estimate.
seTE: Standard error estimate for the log hazard ratio.
n1: The sample size of arm 1.
n2: The sample size of arm 2.
Noma, H. and Maruo, K. (2025). Network meta-analysis combining survival and count outcome data: A simple frequentist approach. medRxiv: 10.1101/2025.01.23.25321051.
Salika, T., Turner, R. M., Fisher, D., Tierney, J. F. and White, I. R. (2022). Implications of analysing time-to-event outcomes as binary in meta-analysis: empirical evidence from the Cochrane Database of Systematic Reviews. BMC Medical Research Methodology 22, 73.
data(woods2) pairwiseHR(treat, studlab=study, event=d, n, data=woods2)data(woods2) pairwiseHR(treat, studlab=study, event=d, n, data=woods2)
A network meta-analysis dataset summarized in hazard ratio statistics provided in Woods et al. (2010).
studlab: ID variable of studies.
treat1: Treatment 1.
treat2: Treatment 2.
TE: Log hazard ratio estimate.
seTE: Standard error estimate of the log hazard ratio estimator.
n1: Sample size 1.
n2: Sample size 2.
data(woods1)data(woods1)
A data frame for network meta-analysis with 2 trials.
Woods, B. S., Hawkins, N. and Scott, D. A. (2010). Network meta-analysis on the log-hazard scale, combining count and hazard ratio statistics accounting for multi-arm trials: A tutorial. BMC Medical Research Methodology 10: 54.
A network meta-analysis dataset reported as dichotomized data provided in Woods et al. (2010).
study: ID variable of studies.
treat: Treatment.
d: The number of events.
n: Sample size.
data(woods2)data(woods2)
A data frame for network meta-analysis with 3 trials.
Woods, B. S., Hawkins, N. and Scott, D. A. (2010). Network meta-analysis on the log-hazard scale, combining count and hazard ratio statistics accounting for multi-arm trials: A tutorial. BMC Medical Research Methodology 10: 54.