Improving accuracy of clinical coding in surgery: collaboration is key

Identification of patients undergoing chronic kidney replacement therapy in primary and secondary care data: validation study based on OpenSAFELY and UK Renal Registry

Objective

To validate primary and secondary care codes in electronic health records to identify people receiving chronic kidney replacement therapy based on gold standard registry data.

Design

Validation study using data from OpenSAFELY and the UK Renal Registry, with the approval of NHS England.

Setting

Primary and secondary care electronic health records from people registered at 45% of general practices in England on 1 January 2020, linked to data from the UK Renal Registry (UKRR) within the OpenSAFELY-TPP platform, part of the NHS England OpenSAFELY covid-19 service.

Participants

38 745 prevalent patients (recorded as receiving kidney replacement therapy on 1 January 2020 in UKRR data, or primary or secondary care data) and 10 730 incident patients (starting kidney replacement therapy during 2020), from a population of 19 million people alive and registered with a general practice in England on 1 January 2020.

Main outcome measures

Sensitivity and positive predictive values of primary and secondary care code lists for identifying prevalent and incident kidney replacement therapy cohorts compared with the gold standard UKRR data on chronic kidney replacement therapy. Agreement across the data sources overall, and by treatment modality (transplantation or dialysis) and personal characteristics.

Results

Primary and secondary care code lists were sensitive for identifying the UKRR prevalent cohort (91.2% (95% confidence interval (CI) 90.8% to 91.6%) and 92.0% (91.6% to 92.4%), respectively), but not the incident cohort (52.3% (50.3% to 54.3%) and 67.9% (66.1% to 69.7%)). Positive predictive values were low (77.7% (77.2% to 78.2%) for primary care data and 64.7% (64.1% to 65.3%) for secondary care data), particularly for chronic dialysis (53.7% (52.9% to 54.5%) for primary care data and 49.1% (48.0% to 50.2%) for secondary care data). Sensitivity decreased with age and index of multiple deprivation in primary care data, but the opposite was true in secondary care data. Agreement was lower in children, with 30% (295/980) featuring in all three datasets. Half (1165/2315) of the incident patients receiving dialysis in UKRR data had a kidney replacement therapy code in the primary care data within three months of the start date of the kidney replacement therapy. No codes existed whose exclusion would substantially improve the positive predictive value without a decrease in sensitivity.

Conclusions

Codes used in primary and secondary care data failed to identify a small proportion of prevalent patients receiving kidney replacement therapy. Codes also identified many patients who were not recipients of chronic kidney replacement therapy in UKRR data, particularly dialysis codes. Linkage with UKRR kidney replacement therapy data facilitated more accurate identification of incident and prevalent kidney replacement therapy cohorts for research into this vulnerable population. Poor coding has implications for any patient care (including eligibility for vaccination, resourcing, and health policy responses in future pandemics) that relies on accurate reporting of kidney replacement therapy in primary and secondary care data.

From coding to clinical nurse specialist: how a review of coding practice enabled hysteroscopy nurse development

Clinical coding, the method by which departments are reimbursed for providing services to patients, is widely mispractised within the NHS. Improving clinical coding accuracy therefore offers an opportunity to increase departmental income, guide efficient resource allocation and enable staff development. The authors audited the clinical coding in outpatient hysteroscopy clinics at their institution and found that coding errors were both prevalent and correctable. By implementing simple changes in coding procedure, and without any additional administrative cost, they significantly improved coding accuracy and achieved an increase in total annual tariffs. Although not applicable in a block contract, this will become highly relevant in a restoration of the Payment by Results tariff system. Nurse development is a key objective of the NHS Long Term Plan but can be hindered by staff costs, which require departmental funding. In the authors' institution, improved clinical coding accuracy directly led to a departmental restructuring, funded the development of a new hysteroscopy nurse development and improved care delivery. Coding errors are not unique to the authors' trust, yet simple amendments led to meaningful changes. Therefore, careful auditing and implemented change are needed to raise national clinical coding standards, to enable clinical restructuring, staff development, and provide more efficient, patient-centred care.

Moving toward a standardized diagnostic statement of pituitary adenoma using an information extraction model: a real-world study based on electronic medical records

PURPOSE

Diagnostic statements for pituitary adenomas (PAs) are complex and unstandardized. We aimed to determine the most commonly used elements contained in the statements and their combination patterns and variations in real-world clinical practice, with the ultimate goal of promoting standardized diagnostic recording and establishing an efficient element extraction process.

METHODS

Patient medical records from 2012 to 2020 that included PA among the first three diagnoses were included. After manually labeling the elements in the diagnostic texts, we obtained element types and training sets, according to which an information extraction model was constructed based on the word segmentation model "Jieba" to extract information contained in the remaining diagnostic texts.

RESULTS

A total of 576 different diagnostic statements from 4010 texts of 3770 medical records were enrolled in the analysis. The first ten diagnostic elements related to PA were histopathology, tumor location, endocrine status, tumor size, invasiveness, recurrence, diagnostic confirmation, Knosp grade, residual tumor, and refractoriness. The automated extraction model achieved F1-scores that reached 100% for all ten elements in the second round and 97.3-100.0% in the test set consisting of an additional 532 diagnostic texts. Tumor location, endocrine status, histopathology, and tumor size were the most commonly used elements, and diagnoses composed of the above elements were the most frequent. Endocrine status had the greatest expression variability, followed by Knosp grade. Among all the terms, the percentage of loss of tumor size was among the highest (21%). Among statements where the principal diagnoses were PAs, 18.6% did not have information on tumor size, while for those with other diagnoses, this percentage rose to 48% (P < 0.001).

CONCLUSION

Standardization of the diagnostic statement for PAs is unsatisfactory in real-world clinical practice. This study could help standardize a structured pattern for PA diagnosis and establish a foundation for research-friendly, high-quality clinical information extraction.

Demystifying the role of anaesthetists in clinical coding in the Australian healthcare system

Despite the self-evident importance of hospital funding, many anaesthetists remain unsure of exactly how their daily work relates to hospital reimbursement. A lack of awareness of the nuances of the Australian hospital activity-based funding system has the potential to affect anaesthetic department reimbursement and thus resourcing. Activity-based funding relies on clinical coders reviewing clinical documentation and quantifying the care given to a patient during an admission. Errors in funding allocation may arise when there is a disconnect between the work performed and the information coded. In anaesthesia, there are several factors impeding this process, including clinical understanding of coding, system setup and coders' understanding of anaesthesia. This article explores these factors from the clinical anaesthetist's point of view and suggests solutions, such as awareness and education, clinician-coder cooperation and redesign of documentation systems at a systems level that anaesthetic departments can incorporate.

Surgeon involvement in clinical coding to improve data accuracy and remuneration in a shoulder and elbow unit

BACKGROUND

Clinical coders are dependent on clear data regarding diagnoses and procedures to generate an accurate representation of clinical activity and ensure appropriate remuneration is received. The accuracy of this process may potentially be improved by collaboration with the surgical team.

METHODS

Between November 2017 and November 2019, 19 meetings took place between the Senior Clinical Fellow of our tertiary Shoulder & Elbow Unit and the coding validation lead of our Trust. At each meeting, the Clinical Fellow assessed the operative note of cases in which uncertainty existed as to the most suitable clinical codes to apply and selected the codes which most accurately represented the operative intervention performed.

RESULTS

Over a 24-month period, clinical coding was reviewed in 153 cases (range 3-14 per meeting, mean 8). Following review, the clinical coding was amended in 102 (67%) of these cases. A total of £115,160 additional income was generated as a result of this process (range £1677-£15,796 per meeting, mean £6061). Only 6 out of 28 (21%) cases initially coded as arthroscopic sub-acromial decompressions were correctly coded as such.

DISCUSSION

Surgeon input into clinical coding greatly improves data quality and increases remuneration received for operative interventions performed.

Automated ICD coding for primary diagnosis via clinically interpretable machine learning

BACKGROUND

Computer-assisted clinical coding (CAC) based on automated coding algorithms has been expected to improve the International Classification of Disease, tenth version (ICD-10) coding quality and productivity, whereas studies oriented to primary diagnosis auto-coding are limited in the Chinese context.

OBJECTIVE

This study aims at developing a machine learning (ML) model for automated primary diagnosis ICD-10 coding.

METHODS

A total of 71,709 admissions in Fuwai hospital were included to carry out this study, corresponding to 168 primary diagnosis ICD-10 codes. Based on clinical implications, two feature engineering methods were used to process discharge diagnosis and procedure texts into sequential features and sequential grouping features respectively by which two kinds of models were built and compared. One baseline model using one-hot encoding features was considered. Light Gradient Boosting Machine (LightGBM) was adopted as the classifier, and grid search and cross-validation were used to select the optimal hyperparameters. SHapley Additive exPlanations (SHAP) values were applied to give the interpretability of models.

RESULTS

Our best prediction model was developed based on sequential grouping features. It showed good performance in the test phase with accuracy and macro-averaged F1 (Macro-F1) of 95.2% and 88.3% respectively. The comparison of the models demonstrated the effectiveness of the sequential information and the grouping strategy in boosting model performance (P-value < 0.01). Subgroup analysis of the best model on each individual code manifested that 91.1% of the codes achieved the F1 over 70.0%.

CONCLUSIONS

Our model has been demonstrated its effectiveness for automated primary diagnosis coding in the Chinese context and its results are interpretable. Hence, it has the potential to assist clinical coders to improve coding efficiency and quality in Chinese inpatient settings.

Causes of death after emergency general surgical admission: population cohort study of mortality

Background

A substantial number of patients treated in emergency general surgery (EGS) services die within a year of discharge. The aim of this study was to analyse causes of death and their relationship to discharge diagnoses, in patients who died within 1 year of discharge from an EGS service in Scotland.

Methods

This was a population cohort study of all patients with an EGS admission in Scotland, UK, in the year before death. Patients admitted to EGS services between January 2008 and December 2017 were included. Data regarding patient admissions were obtained from the Information Services Division in Scotland, and cross-referenced to death certificate data, obtained from the National Records of Scotland.

Results

Of 507 308 patients admitted to EGS services, 7917 died while in hospital, and 52 094 within 1 year of discharge. For the latter, the median survival time was 67 (i.q.r. 21–168) days after EGS discharge. Malignancy accounted for 48 per cent of deaths and was the predominant cause of death in patients aged over 35 years. The cause of death was directly related to the discharge diagnosis in 56.5 per cent of patients. Symptom-based discharge diagnoses were often associated with a malignancy not diagnosed on admission.

Conclusion

When analysed by subsequent cause of death, EGS is a cancer-based specialty. Adequate follow-up and close links with oncology and palliative care services merit development.

Definitions and Prevalence of Multimorbidity in Large Database Studies: A Scoping Review

Background: Multimorbidity presents a key challenge to healthcare systems globally. However, heterogeneity in the definition of multimorbidity and design of epidemiological studies results in difficulty in comparing multimorbidity studies. This scoping review aimed to describe multimorbidity prevalence in studies using large datasets and report the differences in multimorbidity definition and study design. Methods: We conducted a systematic search of MEDLINE, EMBASE, and CINAHL databases to identify large epidemiological studies on multimorbidity. We used the Preferred Reporting Items for Systematic Reviews and Meta-analysis Extension for Scoping Reviews (PRISMA-ScR) protocol for reporting the results. Results: Twenty articles were identified. We found two key definitions of multimorbidity: at least two (MM2+) or at least three (MM3+) chronic conditions. The prevalence of multimorbidity MM2+ ranged from 15.3% to 93.1%, and 11.8% to 89.7% in MM3+. The number of chronic conditions used by the articles ranged from 15 to 147, which were organized into 21 body system categories. There were seventeen cross-sectional studies and three retrospective cohort studies, and four diagnosis coding systems were used. Conclusions: We found a wide range in reported prevalence, definition, and conduct of multimorbidity studies. Obtaining consensus in these areas will facilitate better understanding of the magnitude and epidemiology of multimorbidity.

Validation of the Preoperative Score to Predict Postoperative Mortality (POSPOM) in Germany

BACKGROUND

The Preoperative Score to Predict Postoperative Mortality (POSPOM) based on preoperatively available data was presented by Le Manach et al. in 2016. This prognostic model considers the kind of surgical procedure, patients' age and 15 defined comorbidities to predict the risk of postoperative in-hospital mortality. Objective of the present study was to validate POSPOM for the German healthcare coding system (G-POSPOM).

METHODS AND FINDINGS

All cases involving anaesthesia performed at the University Hospital Bonn between 2006 and 2017 were analysed retrospectively. Procedures codified according to the French Groupes Homogènes de Malades (GHM) were translated and adapted to the German Operationen- und Prozedurenschlüssel (OPS). Comorbidities were identified by the documented International Statistical Classification of Diseases (ICD-10) coding. POSPOM was calculated for the analysed patient collective using these data according to the method described by Le Manach et al. Performance of thereby adapted POSPOM was tested using c-statistic, Brier score and a calibration plot. Validation was performed using data from 199,780 surgical cases. With a mean age of 56.33 years (SD 18.59) and a proportion of 49.24% females, the overall cohort had a mean POSPOM value of 18.18 (SD 8.11). There were 4,066 in-hospital deaths, corresponding to an in-hospital mortality rate of 2.04% (95% CI 1.97 to 2.09%) in our sample. POSPOM showed a good performance with a c-statistic of 0.771 and a Brier score of 0.021.

CONCLUSIONS

After adapting POSPOM to the German coding system, we were able to validate the score using patient data of a German university hospital. According to previous demonstration for French patient cohorts, we observed a good correlation of POSPOM with in-hospital mortality. Therefore, further adjustments of POSPOM considering also multicentre and transnational validation should be pursued based on this proof of concept.

Improving the quality of clinical coding and payments through student doctor-coder collaboration in a tertiary haematology department

Hospitals within the UK are paid for services provided by 'Payment-by-Results'. In a system that rewards productivity, effective collaboration between coders and clinicians is crucial. However, clinical coding is frequently error prone and has been shown to impact negatively on departmental revenue. Our aim was to increase the median number of diagnostic codes per sickle cell inpatient admission at Guy's Hospital by 3. Three interventions were implemented using the Plan, Do, Study, Act structure. This consisted of student doctors searching for diagnoses along with comorbidities that clinical coders had missed, distributing laminated cards with common clinical codes and implementing discharge pro formas. Through auditing, student doctors generated a total of £58 813 over 16 weeks. We observed an increase in the median number of codes by ≥2 additional codes. We improved coding accuracy where we identified errors in an average of 32.5% of admissions each month, improving the quality of patient documentation. We have demonstrated student doctor involvement in clinical coding as a potentially sustainable means of achieving accurate payment for services provided; increasing departmental revenue. We are the first to report the efficacy of student-coder collaboration in improving the accuracy of clinical coding.

Cholangiocarcinoma miscoding in hepatobiliary centres

INTRODUCTION

Cholangiocarcinoma (CCA) are sub-divided into intrahepatic (iCCA) or extrahepatic (eCCA). eCCA are further subdivided into perihilar (pCCA) and distal (dCCA). Current and previous versions of the WHO International Coding of Disease and Oncology classifications (ICD) have separate topography codes for iCCA and eCCA, but none for pCCA. Over recent decades, multiple studies report rising incidence rates of iCCA with declining rates of eCCA, without reference to pCCA. We hypothesised the lack of a specific code for pCCA has led to errors CCA coding, specifically with miscoding of pCCA as iCCA.

METHODS

Clinical notes of cases coded as hepatobiliary carcinoma using ICD-10 criteria (C22.1/Intrahepatic Bile Duct carcinoma, C24.0/Extrahepatic Bile Duct carcinoma, C23X/Malignant Neoplasm Gall Bladder, C22.0/Malignant Neoplasm Liver Cell Carcinoma) over a 2 year period (2015-2017), were reviewed by two independent clinicians at three independent UK regional HepatoPancreatoBiliary centres. The agreed final diagnosis was compared to the originally allocated ICD-10 code.

RESULTS

Of the 625 CCA cases fully reviewed, 226 were coded as C22.1/iCCA. 98 (43%) of these were true iCCA and coded correctly, while 76 cases (34%) were actually pCCA. 92% all pCCA cases were incorrectly coded as iCCA.

CONCLUSION

CCA coding misclassification in UK HPB centres is common, particularly the miscoding of pCCA, which is extrahepatic and the commonest form of CCA, as iCCA. This may be contributing to apparent rising incidence rates of iCCA. Our findings confirm the need to implement distinct topographical codes for iCCA, pCCA and dCCA in future iterations of ICD.

Development and validation of an algorithm of diagnostic and procedural codes for the identification of children hospitalized with a tracheostomy in Ontario, Canada

BACKGROUND

The requirement for a tracheostomy in children is associated with significant morbidity, mortality, and healthcare utilization. Easy identification of children with tracheostomies would facilitate important research on this population and provide quality improvement initiatives.

AIM

The purpose of this study is to determine whether an algorithm of diagnostic and procedural codes can accurately identify children hospitalized with a tracheostomy using routinely collected health data.

METHODS

Chart reviews were performed at the Children's Hospital of Eastern Ontario (CHEO) and the London Health Sciences Center (LHSC) to establish a true positive cohort of pediatric patients with tracheostomies admitted between 2008 and 2016. A multidisciplinary team developed algorithms of diagnostic and procedural codes contained within the Canadian Institute for Health Information Discharge Abstract Database. Algorithms were tested and refined against the true-positive and true-negative cohort. The accuracy of the diagnostic codes related to tracheostomy complications was also evaluated.

RESULTS

A chart review identified 158 unique children with tracheostomies (77 at CHEO, 81 at LHSC) with 901 individual admissions (401 at CHEO, 507 at LHSC). The best algorithms for identifying children with a tracheostomy had a sensitivity and specificity of more than 99%, a positive predictive value (PPV) of 94.0% and negative predictive value (NPV) of 100%. The algorithm for the identification of tracheostomy-related complications had a sensitivity of 76.7%, a specificity of 65%, PPV of 52.3%, and an NPV of 84.7%.

CONCLUSIONS

This study provides an algorithm for the accurate identification of children hospitalized in Canada with a tracheostomy, facilitating population-level epidemiological research and quality improvement initiatives.

Mortality and Rate of Hospitalization in a Colonoscopy Screening Program From a Randomized Health Services Study

BACKGROUND & AIMS

It is difficult to quantify adverse events related to screening colonoscopy due to lack of valid and adequately powered comparison groups. We compared mortality and rate of unplanned hospitalizations among subjects who underwent screening colonoscopies within the Polish Colonoscopy Screening Program (PCSP) vs unscreened matched controls in Poland.

METHODS

Persons 55-64 years old living in the area covered by the PCSP from 2012 through 2015 were assigned in a (1:1) to a group invited for screening colonoscopy (n = 338,477) or a matched group that would be invited 5 years later (controls, n = 338,557). All subjects in the screening group were assigned proposed screening colonoscopy dates (actual dates when invitees confirmed or rescheduled colonoscopy) and those in the control group were assigned virtual dates corresponding to the matched individuals from the screening group. In the screening group, 55,390 subjects (16.4%) underwent screening colonoscopy. Mortality and hospitalization data were obtained from National Registries. We compared mortality and rate of hospitalization between the groups for defined intervals before and after colonoscopy date. Hospitalizations were divided into related and unrelated to colonoscopy based on ICD codes by 3 specialists. Our primary aim was to compare mortality and hospitalization 6 weeks before and 30 days following the actual or virtual date of colonoscopy in the screening or control group.

RESULTS

In the intent to treat analysis, overall there were no significant differences in mortality between the colonoscopy group and control group (0.22% vs 0.22%; risk difference less than .01%; 95% CI, decrease of 0.02% to 0.02%; P = .913). The overall rate of unplanned hospitalization was significantly higher for the colonoscopy group (2.39% vs 2.31% for the control group; risk difference, 0.08%; 95% CI, 0.01%-0.15%; P=.026) for the entire observation period. This was due to the higher rate of hospitalizations after screening (1.10% vs 1.01% for the control group; risk difference, 0.09%; 95% CI, 0.04%-0.14%; P < .001) including higher proportion of hospitalizations that were assessed as related to colonoscopy (0.24% vs 0.22% for the control group; risk difference, 0.02%; 95% CI, 0.00%-0.05%; P = .046). In the per-protocol analysis, the overall rate of hospitalizations did not differ significantly between control and screening colonoscopy groups (1.87% vs 1.90%; P=.709). However, screening colonoscopy did increase rates of related hospitalizations after the date of screening (from 0.14% to 0.31%; P < .001).

CONCLUSIONS

In an analysis of data from the PCSP, we found high-quality evidence that colonoscopy as a screening intervention does not increase mortality before or after colonoscopy. However, it may be associated with a small but significant increase in unplanned hospitalizations, especially after the colonoscopy is completed.

Evaluation of Use of Technologies to Facilitate Medical Chart Review

INTRODUCTION

While medical chart review remains the gold standard to validate health conditions or events identified in administrative claims and electronic health record databases, it is time consuming, expensive and can involve subjective decisions.

AIM

The aim of this study was to describe the landscape of technology-enhanced approaches that could be used to facilitate medical chart review within and across distributed data networks.

METHOD

We conducted a semi-structured survey regarding processes for medical chart review with organizations that either routinely do medical chart review or use technologies that could facilitate chart review.

RESULTS

Fifteen out of 17 interviewed organizations used optical character recognition (OCR) or natural language processing (NLP) in their chart review process. None used handwriting recognition software. While these organizations found OCR and NLP to be useful for expediting extraction of useful information from medical charts, they also mentioned several challenges. Quality of medical scans can be variable, interfering with the accuracy of OCR. Additionally, linguistic complexity in medical notes and heterogeneity in reporting templates used by different healthcare systems can reduce the transportability of NLP-based algorithms to diverse healthcare settings.

CONCLUSION

New technologies including OCR and NLP are currently in use by various organizations involved in medical chart review. While technology-enhanced approaches could scale up capacity to validate key variables and make information about important clinical variables from medical records more generally available for research purposes, they often require considerable customization when employed in a distributed data environment with multiple, diverse healthcare settings.

Long-term survival following successful abdominal aortic aneurysm repair evaluated using Australian administrative data

BACKGROUND

Long-term survival (LTS) following abdominal aortic aneurysm (AAA) surgery is an outcome that can compare open surgical repair (OSR) and endovascular AAA repair (EVAR) methods. We examined the LTS of persons following successful AAA repair using administrative health data covering the Australian Pharmaceutical Benefits and Medicare Benefits Schemes from 1993 to 2014.

METHODS

Participants undergoing AAA surgery were identified using procedure codes and the last service provision date used as a proxy mortality marker. LTS and relative survival with control populations in those who survived the initial post-operative period were used to compare OSR and EVAR and estimates between the first and second halves of the study.

RESULTS

A total of 2060 persons who had undergone AAA repair were identified. Overall median LTS (95% CI) following elective, ruptured OSR and EVAR were 10.4 (9.1-11.0), 8.5 (6.7-10.3) and 9.7 (8.1-11.3) years, respectively. Relative survival rates at 5 and 10 years were 0.89 and 0.7 for OSR and 0.87 and 0.66 for EVAR. LTS rates were similar for OSR and EVAR in age groups 65-84 years (EVAR/OSR range 0.96-1.16); however, EVAR was superior to OSR in persons aged >85 years at 5 years (EVAR/OSR 1.32, log-rank P < 0.05). Relative survival following all techniques of AAA repair showed no significant change over the duration of the study.

CONCLUSION

LTS following AAA repair was heterogeneous in comparison with control populations and varied with age and procedure. The 5-year LTS following EVAR in persons aged >85 years is superior to OSR. Administrative data can define long-term outcomes following aortic aneurysm surgery and may complement data already collected by surgeons.

Validity of discharge ICD-10 codes in detecting the etiologies of endogenous Cushing's syndrome

OBJECTIVE

To investigate the validity of discharge ICD-10 codes in detecting the etiology of endogenous Cushing's syndrome (CS) in hospitalized patients.

METHODS

We evaluated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of CS etiology-related ICD-10 codes or code combinations by comparing hospital discharge administrative data (DAD) with established diagnoses from medical records.

RESULTS

Coding for patients with adrenocortical adenoma (ACA) and those with bilateral macronodular adrenal hyperplasia (BMAH) demonstrated disappointingly low sensitivity at 78.8% (95% CI: 70.1-85.6%) and 83.9% (95% CI: 65.5-93.9%), respectively. BMAH had the lowest PPV of 74.3% (95% CI: 56.4-86.9%). In confirmed ACA patients, the sensitivity for ACA code combinations was higher in patients initially admitted to the Department of Endocrinology before surgery than that in patients directly admitted to the Department of Urology (90.0 vs 73.1%, P = 0.033). The same phenomenon was observed in the PPV for the BMAH code (100.0 vs 60.9%, P = 0.012). Misinterpreted or confusing situations caused by coders (68.1%) and by the omission or denormalized documentation of symptomatic diagnosis by clinicians (26.1%) accounted for the main source of coding errors.

CONCLUSIONS

Hospital DAD is an effective data source for evaluating the etiology of CS but not ACA and BMAH. Improving surgeons' documentation, especially in the delineation of symptomatic and locative diagnoses in discharge abstracts; department- or disease-specific training for coders and more multidisciplinary collaboration are ways to enhance the applicability of administrative data for CS etiologies.

Factors associated with high-cost hospitalization for peritonitis in children receiving chronic peritoneal dialysis in the United States

BACKGROUND

Although peritonitis causes significant morbidity and mortality in children receiving chronic peritoneal dialysis (CPD), little is known about costs associated with treatment.

METHODS

We analyzed 246 peritonitis-related hospitalizations in the USA, linked by the Standardized Care to Improve Outcomes in Pediatric End Stage Renal Disease (SCOPE) and Pediatric Health Information Systems (PHIS) databases. Multivariable logistic regression was used to assess the relationship between high-cost hospitalizations (at or above the 75th percentile) and patient characteristics. Multivariable modeling was used to assess differences in the service-line specific geometric mean between (1) high- and low-cost (below the 75th percentile) hospitalizations and (2) fungal versus other types of peritonitis. Wage-adjusted hospitalization charges were converted to estimated costs using reported cost-to-charge ratios to estimate the cost of hospitalization.

RESULTS

High-cost hospitalizations were associated with the following: age 3-12 years, Hispanic ethnicity, intensive care unit (ICU) stay, length of stay (LOS), and fungal peritonitis. Whereas absolute standardized cost by service line was significantly different when comparing high- and low-cost hospitalizations, the percentage of total cost by service line was similar in the two groups. Cost per case for fungal peritonitis was higher (p < 0.001) in every service line except pharmacy when compared to other peritonitis cases. The median (IQR) cost of hospitalization for the treatment of peritonitis was $13,655 ($7871, $28434) USD.

CONCLUSIONS

Hospitalization-related costs for peritonitis treatment are substantial and arise from a variety of service lines. Fungal peritonitis is associated with high-cost hospitalization.

Differentiating abdominal procedures in pediatric surgery: The inadequacy of current procedural terminology codes

INTRODUCTION

The ability to use detailed, accurate current procedural terminology (CPT) codes is a key component of effective research. We examined the effectiveness of CPT codes to accurately reflect care in patients undergoing surgery for necrotizing enterocolitis (NEC).

METHODS

A multicenter retrospective analysis of operations on patients with NEC was conducted across 4 institutions between 2011 and 2016. Correlation between operative dictation and CPT coding was analyzed.

RESULTS

A total of 124 patients with NEC diagnosis undergoing exploratory abdominal operations were identified. NEC was improperly diagnosed in 25 patients, who were excluded from further analysis. Of the 99 patients reviewed, the initial exploratory abdominal operation was coded inaccurately in 58 cases (59%). Within these, 15 (26%) had multiple coding errors such that the nature of the original operation was not discernable from the applied codes. Inaccurate codes often did not describe the presence of a mucous fistula (n=27, 44%), ostomy (n=24, 39%), or extra segments of bowel resected (n=9, 16%). The length of bowel resected is not currently described by any CPT codes.

CONCLUSION

CPT coding for abdominal operations does not sufficiently reflect complexity of pediatric surgeries. This study highlights the significance of this inadequacy and its implications in future database studies in the era of electronic medical records.

LEVEL OF EVIDENCE

Level IV.

TYPE OF STUDY

Clinical research study.

Diagnostic accuracy of undernutrition codes in hospital administrative discharge database: improvements needed

OBJECTIVES

Hospital administrative databases are widely used for disease monitoring. Undernutrition is highly prevalent among hospitalized patients but the diagnostic accuracy of undernutrition coding in administrative data is poorly known. This study examined the diagnostic accuracy of undernutrition coding in administrative hospital discharge databases.

METHODS

A retrospective cross-sectional study was conducted using 2013 and 2014 administrative data of the Internal Medicine Unit of the Lausanne University Hospital (n = 2509). Two reference diagnoses were defined: Confirmed undernutrition (2002 nutrition risk screening [NRS-2002] score ≥3 plus body mass index [BMI] < 18.5 kg/m²) and probable undernutrition (NRS-2002 ≥ 3 plus any prescribed nutritional management plus BMI ≥18.5 and <20 kg/m² if age <70 y [ < 22 kg/m² if age ≥70 y]). Missing BMI values were imputed.

RESULTS

Of the 2509 eligible patients, 262 (10.4%) were classified as confirmed and 631 (25.2%) as probable undernutrition. The sensitivity, specificity, and negative and positive predictive values (and corresponding 95% confidence intervals) for undernutrition codes using confirmed undernutrition were 43.0 (37.0-49.3), 87.2 (85.8-88.6), 92.9 (91.7-94.0), and 28.2 (23.8-32.8), respectively. The corresponding values using both confirmed and probable undernutrition were 30.0 (27.2-32.9), 93.4 (92.0-94.6), 66.7 (64.7-68.7), and 75.1 (70.6-79.3), respectively. Similar findings were obtained after stratifying for sex or age groups or restricting the analysis to patients with non-missing BMI data.

CONCLUSIONS

The undernutrition codes in hospital discharge data have good specificity but the sensitivity and positive predictive values are low.

Risk of Recurrent Disease and Surgery Following an Admission for Acute Diverticulitis

BACKGROUND

Diverticular disease accounts for significant morbidity and mortality and may take the form of recurrent episodes of acute diverticulitis. The role of elective surgery is not clearly defined.

OBJECTIVE

This study aimed to define the rate of hospital admission for recurrent acute diverticulitis and risk factors associated with recurrence and surgery.

DESIGN

This is a retrospective population-based cohort study.

SETTINGS

National Health Service hospital admissions for acute diverticulitis in England between April 2006 and March 2011 were reviewed.

PATIENTS

Hospital Episode Statistics data identified adult patients with the first episode of acute diverticulitis (index admission), and then identified recurrent admissions and elective or emergency surgery for acute diverticulitis during a minimum follow-up period of 4 years. Exclusion criteria included previous diagnoses of acute diverticulitis, colorectal cancer, or GI bleeding, and prior colectomy or surgery or death during the index admission.

INTERVENTIONS

There were no interventions.

MAIN OUTCOME MEASURES

The primary outcomes measured were recurrent admissions for acute diverticulitis and patients requiring either elective or emergency surgery during the study period.

RESULTS

Some 65,162 patients were identified with the first episode of acute diverticulitis. The rate of hospital admission for recurrent acute diverticulitis was 11.2%. A logistic regression model examined factors associated with recurrent acute diverticulitis and surgery: patient age, female sex, smoking, obesity, comorbidity score >20, dyslipidemia, and complicated acute diverticulitis increased the risk of recurrent acute diverticulitis. There was an inverse relationship between patient age and recurrence. Similar factors were associated with elective and emergency surgery.

LIMITATIONS

The cases of acute diverticulitis required inpatient management and the use of Hospital Episode Statistics, relying on the accuracy of diagnostic coding.

CONCLUSIONS

This is the largest study assessing the rates of hospital admission for recurrent acute diverticulitis. Knowledge of the rate and risk factors for recurrent acute diverticulitis is required to aid discussion and decision making with patients regarding the need and timing of elective surgery. Some factors associated with recurrence are modifiable; therefore, weight reduction and smoking cessation can be championed. See Video Abstract at http://links.lww.com/DCR/A449.

Clinician-directed improvement in the accuracy of hospital clinical coding

'Payment by results' (PbR) remuneration for healthcare services relies on the accurate conversion of diagnoses into Healthcare Resource Group (HRG) codes that are then reimbursed. Inconsistencies in documentation can result in inaccuracies in this process, with consequent implications for measuring activity, disease incidence and organisational performance. The aim of this study was to determine if clinician involvement increases accuracy in the coding of medical cases. Selected records of medical patients admitted to a London NHS trust between November and December 2016 were reviewed by a coding auditor and a clinician. Any changes to the codes and HRG tariff were noted. In total, 123 cases were considered. Changes in code were made on 68 instances, resulting in an overall increase in remuneration of £39,215; an average of £318 per patient. The primary HRG code was changed in 31 cases which accounted for £28,040 of the increase in tariff. In conclusion, clinician involvement can help with documentation ambiguities, thus improving the accuracy of the coding process in a medical setting. Although such collaborative working offers advantages for both the clinician and the coding team, further work is required to investigate the feasibility of this recommendation on a larger scale.

Does a coding sticker for percutaneous nephrolithotomy improve clinical coding accuracy and increase financial remuneration?

Objective:

This study aimed to evaluate whether a coding sticker for percutaneous nephrolithotomy (PCNL), completed by the surgeon after the operation note, improved the accuracy of clinical coding and the financial remuneration for PCNL.

Patients and methods:

A retrospective study was undertaken including all PCNLs performed in a single centre between October 2014 and June 2016. PCNL clinical coding was obtained and applied to yield a Healthcare Resource Group (HRG) code, which was in turn used to calculate the tariff the Trust received for the case. Remuneration and clinical coding accuracy were compared pre- and post-coding sticker introduction.

Results:

Thirty-three cases were included in the study. Eleven patients were reviewed before the introduction of the sticker and 22 after the introduction of the PCNL sticker. Overall mean clinical coding accuracy improved from 65% to 94% after the stickers’ introduction. This resulted in an overall mean increase in remuneration of £501 per case (from £2946 to £3447).

Conclusion:

The implementation of a simple coding sticker for completion after a PCNL improves clinical coding accuracy and increases the financial remuneration.