compare_mv() applies the statistical test described by Bray and Parkin
(2009) to identify registry percentages of morphologically verified cases
(MV%) that differ from comparison registries. Each registry is tested in
turn, using all other registries in the same stratum as its comparison
population.
Arguments
- x
An object with class
qualityorqualitiescontaining the registry identifierregicode, incident case countsfbs, and MV percentagesmv. An optionalmv_casescolumn supplies exact MV case counts.- alpha
A single numeric significance level strictly between 0 and 1. Defaults to
0.05, corresponding to a chi-squared critical value of approximately 3.84 with one degree of freedom.- reference
An optional external reference table. It must contain regicode, fbs, and mv, or mv_cases; registry and cases are accepted as aliases for regicode and fbs. At least two reference rows are needed in each matched stratum to estimate dispersion.
Value
A tibble containing the stratum identifiers, regicode, fbs, the
observed mv, the MV count used in the calculation mv_cases, the number
of comparison registries reference_n, the estimated common proportion
p, the overdispersion estimate phi, the test statistic z2,
p_value, critical_value, the logical flag unusual, and an evaluation
status.
Details
The function accepts either the merged quality data frame returned by
create_quality(..., collapse = TRUE) or the qualities list returned by
create_quality(..., collapse = FALSE). Strata are detected automatically
from identifier columns such as year, sex, and cancer.
By default, each registry is compared with the other registries in its stratum. Supply reference to use an external reference table instead. External rows are matched to each target stratum using the stratum columns shared by x and reference. For comparison registries indexed by \(i = 1, \ldots, n\), let \(y_i\) be the number of morphologically verified cases and \(d_i\) the total number of incident cases. The common MV proportion and overdispersion are estimated as $$\hat{p} = \frac{\sum_i y_i}{\sum_i d_i}$$ and $$\hat{\phi} = \frac{1}{n - 1} \sum_i \frac{(y_i - \hat{p}d_i)^2}{d_i\hat{p}(1 - \hat{p})}.$$ For the registry under test, indexed by \(j\), the statistic is $$Z^2 = \frac{(y_j - \hat{p}d_j)^2} {\hat{\phi}\hat{p}(1 - \hat{p})d_j}$$, which is compared with a chi-squared distribution with one degree of freedom.
When x contains an mv_cases column, it is used as the exact numerator
\(y\). Otherwise, MV case counts are reconstructed as
round(fbs * mv / 100) from the percentage returned by create_quality().
The mvs column is a different quality indicator and is not used in this
test.
A status other than "ok" indicates that the row could not be evaluated.
Possible values are "zero_target_cases",
"insufficient_reference_registries", "zero_reference_cases",
"boundary_reference_proportion", and "zero_reference_dispersion".
An unusual result indicates a statistical difference from the comparison
registries; it does not by itself identify the reason for that difference.
References
Bray, F., & Parkin, D. M. (2009). Evaluation of data quality in the cancer registry: Principles and methods. Part I: Comparability, validity and timeliness. European Journal of Cancer, 45(5), 747-755. doi:10.1016/j.ejca.2008.11.032
Examples
quality_data <- structure(
data.frame(
regicode = c("A", "B", "C", "D"),
year = "2024",
sex = "1",
cancer = "101",
fbs = c(100, 100, 150, 200),
mv = c(95, 70, 80, 75)
),
class = c("quality", "data.frame")
)
compare_mv(quality_data)
#> # A tibble: 4 × 15
#> regicode year sex cancer fbs mv mv_cases reference_n p phi
#> <chr> <chr> <chr> <chr> <dbl> <dbl> <dbl> <int> <dbl> <dbl>
#> 1 A 2024 1 101 100 95 95 3 0.756 1.65
#> 2 B 2024 1 101 100 70 70 3 0.811 8.79
#> 3 C 2024 1 101 150 80 120 3 0.788 11.0
#> 4 D 2024 1 101 200 75 150 3 0.814 10.5
#> # ℹ 5 more variables: z2 <dbl>, p_value <dbl>, critical_value <dbl>,
#> # unusual <lgl>, status <chr>
