Medicare and Medicaid fraud has taken center stage over the last decade due to the sheer magnitude (in expenditures and enrollment) of the programs. Even a small percentage of fraud in terms of total annual expenditures within Medicare or Medicaid can constitute billions of dollars as an expense that is passed on to the taxpayers of the United States. This paper examines methods of quantifying this fraud by using statistical techniques that are consistent with the Centers for Medicare and Medicaid Services guidelines (CMS, 2005). Our interest is in the total overpayment amount of a population of payments made to a given healthcare provider, denoted &tauY, and specifically and in accordance with CMS guidelines, a 90% lower bound for &tauY. The paper considers classical methods of estimation in which the Central Limit Theorem is used, and also newer methods based on the Hypergeometric distribution. It studies the efficiency of \"pennysampling\" (Edwards et al., submitted) wherein the sampling unit is the penny. It also extends the idea of pennysampling to dollar sampling and allows for different sampling units (beneficiary IDs, claims, and claim lines) to be considered.
","abstract_html":"<p> Medicare and Medicaid fraud has taken center stage over the last decade due to the sheer magnitude (in expenditures and enrollment) of the programs. Even a small percentage of fraud in terms of total annual expenditures within Medicare or Medicaid can constitute billions of dollars as an expense that is passed on to the taxpayers of the United States. This paper examines methods of quantifying this fraud by using statistical techniques that are consistent with the Centers for Medicare and Medicaid Services guidelines (CMS, 2005). Our interest is in the total overpayment amount of a population of payments made to a given healthcare provider, denoted &tau<sub>Y</sub>, and specifically and in accordance with CMS guidelines, a 90% lower bound for &tau<sub>Y</sub>. The paper considers classical methods of estimation in which the Central Limit Theorem is used, and also newer methods based on the Hypergeometric distribution. It studies the efficiency of "pennysampling" (Edwards et al., submitted) wherein the sampling unit is the penny. It also extends the idea of pennysampling to dollar sampling and allows for different sampling units (beneficiary IDs, claims, and claim lines) to be considered. </p>","abstract_has_math":false,"creators":["Lawrence, Matthew"],"institution":null,"degree_name":"M.S.","degree_level":"Campus Access Thesis","degree_discipline":"Statistics","degree_department":null,"school":null,"contributors":["Don Edwards"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2011,"date_issued":"2011-01-01T08:00:00Z","date_published":"2011-01-01T08:00:00Z","updated_at":"2026-07-24T04:37:49Z","subjects":["Medicine and Health Sciences","Statistics and Probability","Pennysampling"],"languages":[],"rights":["© 2011, Matthew Lawrence"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://scholarcommons.sc.edu/etd/2677","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Don Edwards"]},{"key":"dc:creator","label":"Author","values":["Lawrence, Matthew"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["1970-01-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Statistics"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Campus Access Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Medicine and Health Sciences","Statistics and Probability","Pennysampling"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["© 2011, Matthew Lawrence"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://scholarcommons.sc.edu/etd/2677"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Medicare and Medicaid fraud has taken center stage over the last decade due to the sheer magnitude (in expenditures and enrollment) of the programs. Even a small percentage of fraud in terms of total annual expenditures within Medicare or Medicaid can constitute billions of dollars as an expense that is passed on to the taxpayers of the United States. This paper examines methods of quantifying this fraud by using statistical techniques that are consistent with the Centers for Medicare and Medicaid Services guidelines (CMS, 2005). Our interest is in the total overpayment amount of a population of payments made to a given healthcare provider, denoted &tauY, and specifically and in accordance with CMS guidelines, a 90% lower bound for &tauY. The paper considers classical methods of estimation in which the Central Limit Theorem is used, and also newer methods based on the Hypergeometric distribution. It studies the efficiency of \"pennysampling\" (Edwards et al., submitted) wherein the sampling unit is the penny. It also extends the idea of pennysampling to dollar sampling and allows for different sampling units (beneficiary IDs, claims, and claim lines) to be considered.
"]},{"key":"dc:title","label":"Title","values":["Pennysampling: Synopses, Applications, and Extensions"]}]}],"canonical_facts":{"dc:contributor":["Don Edwards"],"dc:creator":["Lawrence, Matthew"],"dc:date.available":["1970-01-01T08:00:00Z"],"dc:description.abstract":["Medicare and Medicaid fraud has taken center stage over the last decade due to the sheer magnitude (in expenditures and enrollment) of the programs. Even a small percentage of fraud in terms of total annual expenditures within Medicare or Medicaid can constitute billions of dollars as an expense that is passed on to the taxpayers of the United States. This paper examines methods of quantifying this fraud by using statistical techniques that are consistent with the Centers for Medicare and Medicaid Services guidelines (CMS, 2005). Our interest is in the total overpayment amount of a population of payments made to a given healthcare provider, denoted &tauY, and specifically and in accordance with CMS guidelines, a 90% lower bound for &tauY. The paper considers classical methods of estimation in which the Central Limit Theorem is used, and also newer methods based on the Hypergeometric distribution. It studies the efficiency of \"pennysampling\" (Edwards et al., submitted) wherein the sampling unit is the penny. It also extends the idea of pennysampling to dollar sampling and allows for different sampling units (beneficiary IDs, claims, and claim lines) to be considered.
"],"dc:identifier":["https://scholarcommons.sc.edu/etd/2677"],"dc:rights":["© 2011, Matthew Lawrence"],"dc:subject":["Medicine and Health Sciences","Statistics and Probability","Pennysampling"],"dc:title":["Pennysampling: Synopses, Applications, and Extensions"],"thesis:degree_discipline":["Statistics"],"thesis:degree_level":["Campus Access Thesis"],"thesis:degree_name":["M.S."]},"updated_at":"2026-07-24T04:37:49Z"}