Abstract
Introduction: Accurate and ethical coding is challenging and directly impacts on Payment by Results (PbR). The aims & objectives of this study were to review the existing pattern of coding for spinal surgery and ascertain its appropriateness & accuracy for surgical procedures, medical co-morbidities and post-op complications.
Methods: A retrospective review of 70 consecutive cervical and 100 consecutive lumbar spine patients who were operated from April 2006 onwards was conducted. The excel sheet provided by coding department, hospital notes – clinic letters, physicians’ entries, theatre notes and laboratory reports (biochemistry/microbiology/histology) – were reviewed. Of the 170 cases, 165 were available for analysis.
Results: Coding data of 5 patients who underwent cervical spine surgeries were not available. Of the 165 cases, the accuracy of primary procedural codes was 93.9% (90.8% cervical & 96% lumbar). However this reduced to 77.6% (75.4% cervical & 79% lumbar) when the accuracy for entire description of performed surgery was considered. The procedural codes did not specifically reflect the surgery performed and lacked reproducibility. Surgical levels were coded incorrectly in 9% of the cases. Cervical surgeries were coded as lumbar in 4 and posterior surgery as anterior in 3 cases respectively. Harvest of iliac crest bone graft was not coded in 5 cases. Medical comorbidities were coded appropriately in 64.2% of the patients (55% cervical & 70% lumbar). The commonly missed comorbidities were drug allergies, hypercholesterolemia, smoking and alcoholism. Post-op adverse events were coded in 75% of the cases (16/20 cervical & 5/8 lumbar). The accuracy was better for lumbar as compared to cervical spinal surgeries.
Conclusion: Coding is a universal language of communication amongst healthcare professionals. Its accuracy is important not just for PbR, but for data quality, audit and research purposes too. The financial implications regarding PbR governed by HRG codes (dictated by OPCS 4.4 & ICD–10 codes) are discussed. The awareness of clinical coding is low amongst junior doctors. Following this study, a clinical coding facilitation form has been introduced to improve data quality. Our plan is to close the audit loop and re-evaluate. Literature emphasises qualification of coders, legible documentation by physicians and interaction between coders and clinicians.
Correspondence should be addressed to Sue Woordward, Britspine Secretariat, 9 Linsdale Gardens, Gedling, Nottingham NG4 4GY, England. Email: sue.britspine@hotmail.com
