P David Cooper1, David R Smart2. 1. Department of Diving and Hyperbaric Medicine, Royal Hobart Hospital, GPO Box 1061L, Hobart, Tasmania 7001, Australia, david.cooper@dhhs.tas.gov.au. 2. Department of Diving and Hyperbaric Medicine, Royal Hobart Hospital, Hobart, Tasmania, Australia.
Abstract
INTRODUCTION: Recent Australian attempts to facilitate disinvestment in healthcare, by identifying instances of 'inappropriate' care from large Government datasets, are subject to significant methodological flaws. Amongst other criticisms has been the fact that the Government datasets utilized for this purpose correlate poorly with datasets collected by relevant professional bodies. Government data derive from official hospital coding, collected retrospectively by clerical personnel, whilst professional body data derive from unit-specific databases, collected contemporaneously with care by clinical personnel. AIM: Assessment of accuracy of official hospital coding data for hyperbaric services in a tertiary referral hospital. METHODS: All official hyperbaric-relevant coding data submitted to the relevant Australian Government agencies by the Royal Hobart Hospital, Tasmania, Australia for financial year 2010-2011 were reviewed and compared against actual hyperbaric unit activity as determined by reference to original source documents. RESULTS: Hospital coding data contained one or more errors in diagnoses and/or procedures in 70% of patients treated with hyperbaric oxygen that year. Multiple discrete error types were identified, including (but not limited to): missing patients; missing treatments; 'additional' treatments; 'additional' patients; incorrect procedure codes and incorrect diagnostic codes. Incidental observations of errors in surgical, anaesthetic and intensive care coding within this cohort suggest that the problems are not restricted to the specialty of hyperbaric medicine alone. Publications from other centres indicate that these problems are not unique to this institution or State. CONCLUSIONS: Current Government datasets are irretrievably compromised and not fit for purpose. Attempting to inform the healthcare policy debate by reference to these datasets is inappropriate. Urgent clinical engagement with hospital coding departments is warranted.
INTRODUCTION: Recent Australian attempts to facilitate disinvestment in healthcare, by identifying instances of 'inappropriate' care from large Government datasets, are subject to significant methodological flaws. Amongst other criticisms has been the fact that the Government datasets utilized for this purpose correlate poorly with datasets collected by relevant professional bodies. Government data derive from official hospital coding, collected retrospectively by clerical personnel, whilst professional body data derive from unit-specific databases, collected contemporaneously with care by clinical personnel. AIM: Assessment of accuracy of official hospital coding data for hyperbaric services in a tertiary referral hospital. METHODS: All official hyperbaric-relevant coding data submitted to the relevant Australian Government agencies by the Royal Hobart Hospital, Tasmania, Australia for financial year 2010-2011 were reviewed and compared against actual hyperbaric unit activity as determined by reference to original source documents. RESULTS: Hospital coding data contained one or more errors in diagnoses and/or procedures in 70% of patients treated with hyperbaric oxygen that year. Multiple discrete error types were identified, including (but not limited to): missing patients; missing treatments; 'additional' treatments; 'additional' patients; incorrect procedure codes and incorrect diagnostic codes. Incidental observations of errors in surgical, anaesthetic and intensive care coding within this cohort suggest that the problems are not restricted to the specialty of hyperbaric medicine alone. Publications from other centres indicate that these problems are not unique to this institution or State. CONCLUSIONS: Current Government datasets are irretrievably compromised and not fit for purpose. Attempting to inform the healthcare policy debate by reference to these datasets is inappropriate. Urgent clinical engagement with hospital coding departments is warranted.