Cystic fibrosis (CF) is a heterogeneous recessive genetic disorder with pathobiologic features that reflect mutations in the fibrosis transmembrane conductance regulator (CFTR) gene. CF is a hereditary disease which is inherited, chronic and progressive health problem that affects over 70,000 people worldwide which still is one of the most common deadly diseases affecting white populations. There is little study of the disease relationship to different skin colors, gender, population or age. There are different ways of diagnosing CF. Machine learning algorithms have the ability to extract knowledge out of sufficient sample dataset. Both traditional and deep learning algorithms have been applied for the diagnosis of CF and have shown acceptable performances. Explainable AI has shown its importance in supporting justification of the decisions of such models. In this structured literature review, we select machine learning based studies in the diagnosis of CF based on the selection strategy of interest. We systematically select, analyze and compare literature that used machine learning models developed for diagnosis of CF based on the quality metrics: accuracy, cross validation, AUROC, precision, recall, specificity, sensitivity, F1 score and external validity. This systematic review shows that the cystic fibrosis foundation (CFF) recommendation diagnosis method, which is based on \(\beta \) -adrenergic sweat chloride test method, has not been supported with machine learning algorithms. In our future study, we plan to explore options of supporting this method with machine learning algorithms.