MCI + Cancer: Preliminary Results

Introduction

• Objective: Compare Mild Cognitive Impairment (MCI) test score results over time between groups affected by cancer, either before or after their MCI diagnosis.

• We will be looking at both total MoCA scores and MoCA-MIS.

• To make both groups comparable, we will use an stable balancing weighting approach (Zubizarreta, 2015).

---

Stable Balancing Weighting

• We focus on two different divisions:

  1. MCI vs MCI + Cancer, and

  2. MCI vs MCI + Cancer (before) vs MCI + Cancer (after)

• This means that the baseline or control group will be the MCI group.

• All groups are unbalanced a priori:

---

Average covariates by group before SBW

TABLE: @maibennett

- After re-weighting using SBW, groups are now balanced with respect to the MCI group:

---

Average covariates by group after SBW

TABLE: @maibennett

Preliminary Results

MCI vs MCI + Cancer

Analyzing the MoCA and MoCA-MIS results grouped by year:

 

::::::

MCI vs MCI + Cancer Before vs MCI + Cancer After

Analyzing the MoCA and MoCA-MIS results grouped by year:

 

::::::

Survival Bias

• There is potential survival bias between the groups, given probabilities of death:

Number of deaths to date by group

TABLE: @maibennett

- While only 13% of the control MCI group has died, over 23% of the group with MCI + Cancer before the diagnosis passed away, and 36% of the group who got a cancer diagnosis after MCI detection.