R/aci.R

In healthequal: Compute Summary Measures of Health Inequality

#'  Absolute concentration index (ACI)
#'
#'  The absolute concentration index (ACI) is an absolute measure of inequality
#'  that indicates the extent to which an indicator is concentrated among
#'  disadvantaged or advantaged subgroups, on an absolute scale.
#'
#'  ACI can be calculated using disaggregated data and individual-level data.
#'  Subgroups in disaggregated data are weighted according to their population
#'  share, while individuals are weighted by sample weight in the case of data
#'  from surveys.
#'
#'  The calculation of ACI is based on a ranking of the whole population from
#'  the most disadvantaged subgroup (at rank 0) to the most advantaged subgroup
#'  (at rank 1), which is inferred from the ranking and size of the subgroups.
#'  ACI can be calculated as twice the covariance between the health indicator
#'  and the relative rank. Given the relationship between covariance and
#'  ordinary least squares regression, ACI can be obtained from a
#'  regression of a transformation of the health variable of interest on the
#'  relative rank. For more information on this inequality measure see
#'  Schlotheuber (2022) below.
#'
#'  **Interpretation:** ACI is 0 if there is no inequality. The larger the
#'  absolute value of ACI, the higher the level of inequality. Positive values
#'  indicate a concentration of the indicator among advantaged subgroups, and
#'  negative values indicate a concentration of the indicator among
#'  disadvantaged subgroups.
#'
#'  **Type of summary measure:** Complex; absolute; weighted
#'
#'  **Applicability:** Ordered dimension of inequality with more than two
#'  subgroups
#'
#'  **Warning:** The confidence intervals are approximate and might be biased.
#'
#' @param est The indicator estimate. Estimates must be available for all
#' subgroups/individuals (unless force=TRUE).
#' @param subgroup_order The order of subgroups/individuals in an increasing
#' sequence.
#' @param pop For disaggregated data, the number of people within each subgroup.
#' This must be available for all subgroups.
#' @param weight The individual sampling weight, for individual-level data from
#' a survey. This must be available for all individuals.
#' @param psu Primary sampling unit, for individual-level data from a survey.
#' @param strata Strata, for individual-level data from a survey.
#' @param fpc Finite population correction, for individual-level data from a
#' survey where sample size is large relative to population size.
#' @param lmin Minimum limit for bounded indicators
#' (i.e., variables that have a finite upper and/or lower limit).
#' @param lmax Maximum limit for bounded indicators
#' (i.e., variables that have a finite upper and/or lower limit).
#' @param conf.level Confidence level of the interval. Default is 0.95 (95%).
#' @param force TRUE/FALSE statement to force calculation with missing
#' indicator estimate values.
#' @param ... Further arguments passed to or from other methods.
#' @examples
#' # example code
#' data(IndividualSample)
#' head(IndividualSample)
#' with(IndividualSample,
#'      aci(est = sba,
#'          subgroup_order = subgroup_order,
#'          weight = weight,
#'          psu = psu,
#'          strata = strata))
#' # example code
#' data(OrderedSample)
#' head(OrderedSample)
#' with(OrderedSample,
#'      aci(est = estimate,
#'          subgroup_order = subgroup_order,
#'          pop = population))
#' @references Schlotheuber, A, Hosseinpoor, AR. Summary measures of health
#' inequality: A review of existing measures and their application. Int J
#' Environ Res Public Health. 2022;19(6):3697. doi:10.3390/ijerph19063697.
#' @return The estimated ACI value, corresponding estimated standard error,
#'  and confidence interval as a `data.frame`.
#' @importFrom stats binomial gaussian glm predict qnorm quantile rnorm vcov
#' @importFrom utils data
#' @importFrom survey svyglm svymean svyvar
#' @importFrom srvyr as_survey
#' @importFrom dplyr lag group_by select
#' @importFrom rlang .data
#' @export
#' @rdname aci
#'
aci <- function(est,
                subgroup_order,
                pop = NULL,
                weight = NULL,
                psu = NULL,
                strata = NULL,
                fpc = NULL,
                lmin = NULL,
                lmax = NULL,
                conf.level = 0.95,
                force = FALSE,
                ...) {

  # Variable checks
  ## Stop
  if (!force) {
    if (anyNA(est))
      stop('Estimates are missing in some subgroups.
           Specify force=TRUE to allow missing values.')
  } else {
    pop <- pop[!is.na(est)]
    subgroup_order <- subgroup_order[!is.na(est)]
    if (!is.null(psu))
      psu <- psu[!is.na(est)]
    if (!is.null(strata))
      strata <- strata[!is.na(est)]
    if (!is.null(weight))
      weight <- weight[!is.na(est)]
    est <- est[!is.na(est)]
  }
  if (length(est) <= 2) {
    stop('Estimates must be available for more than two subgroups')
  }
  if (!is.null(est)) {
    if (!is.numeric(est))
      stop('Estimates need to be numeric')
  }
  if (!is.null(pop)) {
    if (anyNA(pop)) {
      stop('Population is missing in some subgroups')
    }
    if (!is.numeric(pop)) {
      stop('Population variable needs to be numeric')
    }
    if (all(pop == 0)) {
      stop('Population variable is of size 0 in all subgroups')
    }
  }
  if (is.null(subgroup_order)) {
    stop('Subgroup order variable needs to be declared')
  }
  sorted_order <- sort(subgroup_order)
  if (!is.null(pop) &
      (any(diff(sorted_order) != 1) || any(sorted_order %% 1 != 0))) {
    stop('Subgroup order variable must contain integers in increasing order')
  }
  if (!is.null(weight) & !is.numeric(weight)) {
    stop('Weight variable needs to be numeric')
  }
  if (!is.null(lmin) & is.null(lmax) |
      !is.null(lmax) & is.null(lmin)) {
    stop(
      'The minimum limit (lmin) and maximum limit (lmax) should be declared for
      bounded indicators'
    )
  }
  if (!is.null(lmin) & !is.null(lmax)) {
    if (min(est) < lmin)
      stop('Estimate variable has values outside of the specified limits')
    if (lmin == lmax | lmin > lmax)
      stop('The minimum limit (lmin) should be different and less than the
      maximum limit (lmax)')
  }
  ## Warning
  if (is.null(pop) & is.null(weight)) {
    message('Neither a population variable nor a weight variable has been
            declared')
  }
  if (!is.null(lmin) & !is.null(lmax)) {
    message('Bounded indicator normalisation applied')
  }

  # Options
  options(survey.lonely.psu = "adjust")
  options(survey.adjust.domain.lonely = TRUE)

  # Calculate summary measure

  ## Create pop if NULL
  if (is.null(pop) & is.null(weight)) {
    pop <- rep(1, length(est))
  }
  if (is.null(pop) & !is.null(weight)) {
    pop <- weight
  }

  ## Rank subgroups from the most disadvantaged to the most advantaged
  reorder <- order(subgroup_order)
  pop <- pop[reorder]
  subgroup_order <- subgroup_order[reorder]
  if (!is.null(weight)) {
    weight <- weight[reorder]
    intercept <- 1
  } else {
    intercept <- sqrt(pop)
  }
  if (!is.null(strata))
    strata <- strata[reorder]
  if (!is.null(psu))
    psu <- psu[reorder]
  est <- est[reorder]
  if (!is.null(lmin) & !is.null(lmax)) {
    est <- (est - lmin) / (lmax - lmin)
  }
  sumw <- sum(pop, na.rm = TRUE)
  cumw <- cumsum(pop)
  cumw1 <- dplyr::lag(cumw)
  cumw1[is.na(cumw1)] <- 0

  newdat_aci <- as.data.frame(cbind(est,
                                    pop,
                                    psu,
                                    strata,
                                    weight,
                                    subgroup_order,
                                    sumw,
                                    cumw,
                                    cumw1,
                                    intercept))

  newdat_aci <- newdat_aci %>%
    group_by(subgroup_order) %>%
    mutate(cumwr = max(.data$cumw, na.rm = TRUE),
           cumwr1 = min(.data$cumw1, na.rm = TRUE)) %>%
    ungroup()

  rank <- (newdat_aci$cumwr1 + 0.5 *
             (newdat_aci$cumwr - newdat_aci$cumwr1)) / newdat_aci$sumw
  tmp <- (newdat_aci$pop / newdat_aci$sumw) * ((rank - 0.5) ^ 2)
  sigma1 <- sum(tmp)
  tmp1 <- newdat_aci$pop * newdat_aci$est
  meanlhs <- sum(tmp1)
  meanlhs1 <- meanlhs / newdat_aci$sumw
  lhs <- (sigma1 * 2 * (newdat_aci$est / meanlhs1) * newdat_aci$intercept)
  lhs1 <- lhs * meanlhs1
  rhs <- rank * newdat_aci$intercept

  newdat_aci <- as.data.frame(cbind(newdat_aci,
                                    lhs,
                                    lhs1,
                                    rhs))

  ## Calculate ACI
  if (is.null(weight)) {
    mod <- glm(lhs1 ~ 0 + rhs + intercept,
               family = gaussian,
               data = newdat_aci)
  } else {
    tids <- if (is.null(psu)) {
      ~ 1
    } else {
      ~ psu
    }
    tstrata <- if (is.null(strata)) {
      NULL
    } else {
      ~ strata
    }
    tfpc <- if (is.null(fpc)) {
      NULL
    } else {
      ~ fpc
    }

    newdat_aci_s <- svydesign(ids = tids,
                              probs = NULL,
                              strata = tstrata,
                              weights = ~ weight,
                              fpc = tfpc,
                              data = newdat_aci)

    mod <-  svyglm(lhs1 ~ 0 +  rhs + intercept,
                   design = newdat_aci_s,
                   family = gaussian)
  }

  aci <- mod$coefficients[[1]]

  ## Calculate 95% confidence intervals
  se.formula <- sqrt(diag(vcov(mod)))[[1]]
  cilevel <- 1 - ((1 - conf.level) / 2)
  lowerci <- aci - se.formula * qnorm(cilevel)
  upperci <- aci + se.formula * qnorm(cilevel)

  # Return data frame
  return(data.frame(measure = "aci",
                    estimate = aci,
                    se = se.formula,
                    lowerci = lowerci,
                    upperci = upperci))
}