Data used to model a triage scoring scale for a Danish emergency department.

The Lackey study was a prospective cohort study of adult smear-positive tuberculosis (TB) patients enrolled between January 2010 and December 2011 with no prior TB disease. Data from the cohort was used to model the association of several predictors with a treatment interruption before the complete courses of therapy. The analysis of treatment outcome in original article uses methods for binary data. A time-to-event analysis might be more appropriate but the dataset does not have data sufficient for that analysis.

Usage

tb.interruption

Format

A tibble with 1293 rows and 18 variables:

id

Character vector, unique participant ID

age.group

A factor with 4 levels: 21 and younger; 22 to 26; 27 to 37; 38 and older

bmi

a factor with 3 levels: Normal; Overweight/Obese; Underweight. These categories reflect older WHO coding and do not apply to all populations.

chronic.disease

a factor with two levels: No, no other chronic disease; Yes, other chronic diseases present in the participant

hiv.test

Outcome of HIV test, a factor with 3 levels: Negative; Positive; Test not Done

marital.status

a factor with 4 levels: Divorced/separated; Married/cohabitating; Single; Widowed

poverty

socioeconomic status, a factor with two levels: No, not living in extreme poverty; Yes, living in extreme poverty

prison.history

a factor with 2 levels: No, no history of having been incarcerated; Yes, participant has been incarcerated

education

a factor with 2 levels: No, participant does not have at least a secondary school education; Yes, participant does have a secondary school education

tobacco.use

a factor with 3 levels: Currently smokes; Never smoked; Used to Smoke

alcohol.use

a factor with 2 levels: No, participant does not use alcohol at least weekly; Yes, participant does use alcohol at least weekly

drug.use

a factor with 2 levels: No, history of illicit drug use; Yes, a history of illicit drug use

rehab.history

a factor with 2 levels: No, no history of residence in a rehabilitation facility; Yes, prior residence in a rehabilitation facility

mdr.tb

a factor with two levels: No, participant has not been treated for multi-drug resistant TB; Yes, participant has been treated for MDR TB

diabetes

a factor with 2 levels: No, participant does not have type 2 diabetes; Yes, participant does have diabetes

trt.outcome

a factor with 4 levels denoting treatment outcome: Cured; Default (treatment was interrupted before 2 months); Died; Still in treatment; Transferred out

Source

doi:10.5061/dryad.fp94d

References

Lackey, Brian, et al. "Patient characteristics associated with tuberculosis treatment default: a cohort study in a high-incidence area of Lima, Peru." PLoS One 10.6 (2015): e0128541. doi:10.1371/journal.pone.0128541