Concordance between two models of stratification for patients living with HIV infection to providing pharmaceutical care

Abstract

Objective

To determine the degree of agreement of 2 differents stratification models for pharmaceutical care to people living with HIV.

Methods

This was a single-center observational prospective cohort study of patients with regular follow-up in pharmaceutical care consultations according to the Capacity–Motivation–Opportunity methodology, conducted between January 1 and March 31, 2023.

Patients received the pharmacotherapeutic interventions applied routinely to ambulatory care patients according to this model. As part of the usual clinical practice, the presence or absence of the variables that apply to both stratification models were collected.

The scores obtained and the corresponding stratification level were collected for each patient according to both stratification models published (ST-2017 and ST-2022).

To analyze the reliability between the measurements of 2 numerical score models of the stratification level with both tools, their degree of concordance was calculated using the intraclass correlation coefficient. Likewise, reliability was also evaluated from a qualitative perspective by means of Cohen's Kappa coefficient.

Additionally, the existence of correlation between the scores of the 2 models was assessed by calculating Pearson's correlation coefficient.

Results

Of the total of 758 patients being followed in the cohort, finally, 233 patients were enrolled. The distribution of patients for each stratification model was: ST-2017: 59.7% level-3, 25.3% level-2, and 15.0% level-1, while for ST-2022: 60.9% level-3, 26.6% level-2, and 12.4% level-1.

It was observed that the reclassification was symmetrical (P = .317). The qualitative analysis of the agreement between the models showed a good Cohen's kappa value, (K = 0.66). A value of 0.563 was found as the intraclass correlation coefficient.

Finally, the correlation analysis between the quantitative scores of the 2 models yielded a Pearson correlation coefficient of 0.86.

Conclusions

The concordance between the 2 models was good, which confirms that the multidimensional adaptation and simplification of the model were correct and that its use can be extended in routine clinical practice.