Medical disinformation poses a serious threat to medical demographic security by distorting health behavior at scale, often amplified by confusion among individuals seeking reliable medical information across diverse topics. These distortions can increase vaccine hesitancy, encourage unproven or harmful practices such as ingesting bleach as a purported COVID-19 treatment, and delay evidence-based care. Medical disinformation also erodes trust in health institutions and contributes to cumulative harms, including increased morbidity and mortality, widening health disparities, and, in some cases, real-world violence linked to conspiracy narratives. Despite rapid advances in automated detection methods, the evidence base remains fragmented, obscuring dominant approaches, required resources, and critical research gaps. This paper presents a systematic review of medical disinformation detection research. Major modeling paradigms and reported evaluation evidence are synthesized, encompassing traditional machine learning, deep learning and transformer-based models, knowledge graph approaches, and fact-checking pipelines, together with the datasets and medical knowledge resources that support them. Commonly used feature types are categorized, their strengths and limitations are assessed, persistent weaknesses in resources and detection pipelines are identified, and targeted recommendations are offered to improve future systems and support more reliable medical informatics that strengthens medical demographic security (pp.81-94).