testing_incidence: Incindence from testing history

Description Usage Arguments Details Value

Description

Incindence from testing history

Usage

1
2
3
4
testing_incidence(report_hiv_pos, biomarker_art, low_viral, hiv, last_hiv_lower,
  last_hiv_upper, ever_hiv_test, weights = rep(1,
  length(report_hiv_pos))/length(report_hiv_pos), distribution = "weibull",
  ptruth = NULL, ptreated = NULL, non_tester_tid = 123.824)

Arguments

report_hiv_pos

A logical vector indicating whether each subject reported a positive hiv status

biomarker_art

A logical vector indicating whether ART antibodies are present. NA if test not done.

low_viral

A logical vector indicating whether viral load is <=1000

hiv

A logical vector indicating hiv status

last_hiv_lower

A numeric vector indicating the lower bound of the months since last hiv test bin in months.

last_hiv_upper

A numeric vector indicating the upper bound of the months since last hiv test bin in months.

ever_hiv_test

A logical vector indicating whether the subject had eer had an hiv test.

weights

Survey weights

distribution

Either "weibull" or "gamma." This controls the family of distribution used to model time since last test.

ptruth

The proportion of the diagnosed hiv positive population that would report being hiv positive. If NULL, this is estimated using biomarker_art and low_viral.

ptreated

The proportion of hiv positive individuals with postive biomarker_art or low_viral. If NULL, this is estimated from the data.

non_tester_tid

mean nummber of months between infection and diagnosis for non_testers.

Details

last_hiv_upper should always be greater than last_hiv_lower, and may be infinite (e.g. test was > 24 months ago). Those with missing data or who never had an hiv test should be assigned NA for both their lower and upper bound.

Value

A data.frame of class test_inc with elements: 'inc': the estimated incidence. 'pundiag': The estimated proportion of positive cases that are undiagnosed. 'psay_undiag': The proportion of positive cases that report being undiagnosed. 'pmiss_class': the proportion whose diagnosis status is incorrectly reported by the individual . 'phiv': The proportion with a positive diagnosis. 'ptester': The proportion who have ever been tested. 'mean_time_since_last_test': the mean tie since last test in years. 'tid': mean time between infection and diagnosis in years. 'ptruth' the proportion of positive individuals that correctly report their status. 'ptreated': the proportion of positive indivudals who are identified as treated by viral load or biomarker.


ifellows/TestingHistoryIncidence documentation built on May 30, 2019, 8:04 p.m.