Methods Appendix e-1 - Neurology:CP



Appendix e-1

Study Populations

Using an administrative case definition validated in Manitoba and Nova Scotia,e1,e 2 we identified all persons with MS in each province as those with ≥3 hospital or physician claims for MS (ICD-9/10 = 340/G35).We selected a matched cohort from the general population after excluding individuals with any diagnostic codes (ICD-9/10) for demyelinating disease: optic neuritis [377.3/H46], acute transverse myelitis [323.82/G37], acute disseminated encephalomyelitis [323/G36.9], demyelinating disease of CNS unspecified [341.9/G37.8], other acute disseminated demyelination [G36], MS [340/G35], or neuromyelitis optica [341.0/G36.0].

Comorbidities

In each province we applied case definitions for diabetes, hypertension, hyperlipidemia and IHD that were validated in Manitoba and Nova Scotia (see below).

Table e-1: Administrative case definitions for comorbidity

|Comorbidity |Diagnostic Codes |Administrative Case Definition |

| |ICD -9 codes |ICD -10 codes |No. Years |No. Hospital or Physician Claims |

| | | |of Data | |

|Diabetes |250 |E10-E14 |5 |≥1H or ≥2P |

|Hypertension |401-405 |I10-I13, I15 |2 |≥1H or ≥2P |

|Heart disease |410-414 |I20-I25 |5 |≥1H or ≥2P |

|Hyperlipidemia |272 |E780, E782, E784, E785 |5 |≥1H or ≥2P |

H = hospital, P = physician, ICD = International Classification of Disease

Matched analysis

As this was a matched cohort design, a matched analysis is not needede3 nor is adjustment needed to control for confounding due to the matched variables if follow-up time is the same in both cohorts. If follow-up time is not the same due to differential survival,e3 then adjustment is needed.

Meta-analysis

Unlike fixed effects meta-analysis which assumes a common effect size across sites (provinces), random effects meta-analysis assumes that the effect size varies across sites. In fixed effects meta-analysis the individual study weights are the inverse of the variance, thus giving more weight to effects based on large sample sizes. In random effects meta-analysis the weights are the inverse of a variance that includes two components, the within study (province) variance and between study (province) variance. This has the effect of giving smaller studies (provinces) more weight and larger studies (provinces) less weight than in the fixed effects model.

We assessed homogeneity of the estimates using the I2 index of heterogeneity.e4 I2 describes the proportion of variation in point estimates due to heterogeneity of studies rather than to sampling error and is constructed based on the Q statistic (the weighted sum of squared differences between individual study effects). I2 = [(Q – degrees of freedom)/Q]*100 where values of I2 of 75% are considered high.4 If heterogeneity is high this suggests that there are larger differences than can be explained by chance variation; there is more than one population in the different studies. We report τ2 as a measure of between study variance.

e-References

e1. Marrie RA, Yu N, Blanchard JF, Leung S, Elliott L. The rising prevalence and changing age distribution of multiple sclerosis in Manitoba. Neurology 2010;74:465-471.

e2. Marrie RA, Fisk J, Stadnyk K, et al. The incidence and prevalence of multiple sclerosis in Nova Scotia, Canada Can J Neurol Sci 2013;40:824-831.

e3. Rothman KJ, Greenland S, eds. Modern Epidemiology, 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins, 1998.

e4. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses, 2003.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download