. Author manuscript; available in PMC: 2019 Oct 1.

Published in final edited form as: J R Stat Soc Ser A Stat Soc. 2018 Feb 26;181(4):1193–1209. doi: 10.1111/rssa.12357

Table 4.

Estimated difference in mean change in CD4 cell count^a and corresponding 95% confidence intervals (CIs) in two target populations (all men and women combined and all women living with HIV in the US) based on data from ACTG trials, the WIHS, and CNICS.^b

Cohort

Trial

Difference in Mean Change (95% CI)

{\hat{Δ}}_{T}

{\hat{Δ}}_{S}

{\hat{Δ}}_{IPSW}

WIHS

493

320^c

200

24 (7, 41)

38 (17, 59)

46 (23, 70)

WIHS

1,012

A5202^d

322

1 (−35, 37)

−19 (−62, 25)

35 (−45, 115)

CNICS

6,158

320

1,156

19 (12, 25)

18 (9, 26)

17 (9, 25)

CNICS

12,302

A5202

1,857

6 (−8, 20)

7 (−18, 32)

−2 (−31, 28)

For 320, the outcome was change in CD4 cell count from baseline to week 4. For A5202, the outcome was change in CD4 cell count from baseline to week 48.

T = within trial; S = stratified; ISPW = inverse probability of sampling weighted.

For 320, the treatment contrast was protease inhibitor (X = 1) vs. no protease inhibitor (X = 0).

For A5202, the treatment contrast was abacavir-lamivudine (X = 1) vs. tenofovir disoproxil fumarate-emtricitabine (X = 0) plus efavirenz or ritonavir-boosted atazanavir.