Journal Club for Analysis of Complex Datasets

1
Journal Club for Analysis of Complex Datasets

• Cui Y, Shu X-O, Gao Y-T, Cai H, Tao M-H, Zheng W.
  Assocation of ginseng use with survival and
  quality of life among breast cancer patients.
• Am J Epidemiol 2006 163 645-653.

2
Summary

• Objective evaluate effects of ginseng use on
  breast cancer survival survivors QOL.
• Design cohort study.
• Setting Shanghai Breast Cancer Study.
• Population-based case-control study.
• Subjects women aged 25-64 years w/ newly
  diagnoses breast cancer Aug1996 Mar1998
  (N1455).
• Cases from Shanghai Study.
• Followed through Dec2002.

3
Summary (2)

• In-person interviews conducted 66 days post-dx.
• Exposure ginseng use.
• Type, duration, frequency of use.
• Outcome
• Overall survival.
• Disease-free survival.
• QOL.
• Assessed during in-person follow-up interviews
  for 1065 of 1248 surviving patients.
• Sociodemographic, medical, other CAM info
  collected as well.

4
Analysis

• Overall survival disease-free survival
• Kaplan-Meier plots log-rank tests.
• Proportional hazards regression.
• Adjusted for age, education, income, stage of
  disease, estrogen progesterone receptor status,
  std cancer txs.
• QOL
• Linear regression.
• Adjusted for age, marital status, education,
  income, menopausal status, recurrence of disase,
  time since cancer dx, use of other CAM.

5
Results Survival

• Ginseng use
• 398 (27.4) regular use.
• 1057 (72.6) nonusers (never used before cancer
  dx).
• All ginseng users received 1 mainstream tx.

6
Overall Survival
7
Disease-Free Survival
8
Results Survival
9
Results QOL

• Ginseng use
• Before/after dx
• Never use before or after.
• Use before but not after.
• Use after but not before.
• Use both before after.
• Ever/current use after dx.
• Cumulative use by current users.

10
(No Transcript)
11
Strengths

• Large population-based sample.
• Excellent participation rate.
• All ginseng users received mainstream tx
• provides well-defined sample to evaluate ginseng
  as complementary rather than alternative therapy.
• Differentiate between types of ginseng use
• Red (hot) used for short periods.
• White Asian American (cold) used for longer
  periods.

12
Strengths (2)

• QOL
• Assessed timing of ginseng use (before/after dx).
• Assessed ever/current use
• Ever users quit using ginseng?
• Placebo/healthy user effect for current users?
• Assessed dose-response.
• caution is required in interpreting the
  resultsfindings need to be confirmed in more
  rigorous and randomized clinical trials.

13
Limitations

• Lack of info on ginseng use after dx for patients
  who died before follow-up interview.
• Unable to exclude confounding by other CAM.
• Self-reported data potential recall
  self-report bias.
• Could not evalute methods of ginseng use.

14
Suggestions Survival

• Time-dependent ginseng use (eg, QOL analysis).
• Account for timing of cancer tx.
• Adjustment for other healthy patient factors
  possibly associated with ginseng?
• Trend analyses of categorized continuous measures
  (eg, cumulative ginseng use)?

15
Suggestions QOL

• lack of ginsengs effect on physical well-being
  obsrved in this study may be due to the fact that
  patients had already fully recovered physically
  by the time the survey was conducted.
• Time-dependent analysis of ginseng use physical
  state.
• Authors state need to further study effect of
  ginseng on physical well-being at earlier time
  points.

16
Sensitivity analysis

• Lin DY, Psaty BM, Kronmal RA. Assessing the
  sensitivity of regression results to unmeasured
  confounders in observational studies. Biometrics
  1998 54 948-963.
• Assess effect of unmeasured binary or normal
  confounder.
• For binary confounder, effect is function of
  prevalence effect among E/E- groups.

17
Survival Binary Unmeasured Confounder (U)

• where
• ? log(HR(E)) when U unobserved
• ? log(HR(E)) when U observed
• ?0 logHR(UE-)
• ?1 logHR(UE)
• P0 Pr(UE-)
• P1 Pr(UE)

18
Sensitivity AnalysisOverall Survival, Any
Ginseng, HR(U) 2.0
19
Sensitivity AnalysisDisease-Free Survival, Any
Ginseng, HR(U)2.0