PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
 
CME J Geriatr Med. Author manuscript; available in PMC 2011 October 12.
Published in final edited form as:
CME J Geriatr Med. 2008; 10(1): 34–36.
PMCID: PMC3191527
EMSID: UKMS1223

Coding Geriatric syndromes: How good are we?

Abstract

High quality coding of hospital activity is important because the data is used for resource allocation and measuring performance. There is little information on the quality of coding of admissions of frail older people who have multiple diagnoses, co-morbidities and functional impairment. Presence or absence of four geriatric syndromes and eight medical conditions was noted on case note review (CNR). Discharge summaries (DS) and hospital coding (HC) were reviewed and compared with the CNR. Forty patients had at least one geriatric syndrome noted in the DS; 16 (40.0%) were captured by the HC. Of 57 patients with at least one medical condition noted in the DS, 52 (91.2%) were captured by the HC (p<0.0001 for difference in HC capture rates). We have demonstrated poor capture of information on geriatric syndromes compared to medical conditions in discharge summaries and hospital coding and propose a problem list bookmark approach to improve this.

Keywords: Coding, routine data, geriatric syndromes, frailty

Key points

  • High quality coding of hospital activity is important because the data is used for resource allocation, service planning and measuring performance.
  • There is little information on the quality of coding of admissions of frail older people who have multiple diagnoses, co-morbidities and functional impairment.
  • This study demonstrates poor capture of information on geriatric syndromes compared to medical conditions in discharge summaries and hospital coding and a problem list bookmark approach is proposed to improve this.

Introduction

There is growing interest in the use of coded hospital inpatient data1. Its value in research is well recognised but its contribution to measuring performance at Trust, department and clinician level is a more recent development in the United Kingdom2,3. Performance indicators inform service planning and increasingly the allocation of healthcare resources. For example, the new financial system of Payment by Results aims to support NHS modernisation by paying hospitals for the work they do. An Audit Commission report highlighted the importance of accurate recording of hospital activity4. However 14% of trusts recorded 3% or more of their activity as unclassified (U codes) which are not reimbursed by the Department of Health. Furthermore limitations of routine data such as hospital episode statistics to monitor individual clinician’s performance have been highlighted in a recent Royal College of Physicians iLab report5,6.

Maximising the quality of coding data in terms of accuracy and completeness is therefore a priority7,8. This depends both on full recording of diagnoses, procedures and complications by the clinician writing the discharge summary and correct translation of this information into clinical codes by trained staff (Figure 1). There have been a number of studies investigating the quality of discharge summaries9,10 and clinical coding for specific diagnoses11,12 showing reasonable accuracy but there is little work on the quality of coding of admissions of frail older people whose admissions are characterized by multiple diagnoses, co-morbidity and functional impairment13. The objective of this study was to determine the completeness of discharge summaries and hospital coding with regard to four geriatric syndromes and eight medical diagnoses ascertained by case note review.

Figure 1
The pathway of information transfer from medical notes to hospital coding

Methods

The study was based on an acute Medicine for Older People ward at Southampton University Hospital NHS Trust.. Organisational approval was obtained and written confirmation was provided from the Local Research Ethics Committee that ethics approval was not required. Case note review was carried out on all patients discharged within a specified four month period by a single specialist registrar in geriatric medicine. Notes from admissions where the patient had died were excluded because the format of these discharge summaries was different.

During the case note review, information was collected on the date of admission, date of discharge, date of birth, presence of any one of four geriatric syndromes - falls, mobility impairment, cognitive impairment and incontinence, or any of the eight most common medical conditions ascertained in a pilot study (myocardial infarction, atrial fibrillation, heart failure, stroke, pneumonia, chronic obstructive pulmonary disease, peptic ulcer disease or urinary tract infection). The information was summarised as a problem list and was used as the ‘gold standard’. The number of patients with a geriatric syndrome or medical diagnosis identified by the hospital discharge summary or coding was compared with the number identified in the case note review problem list.

Statistical analysis

The data were summarised using medians and inter-quartile ranges and frequencies and percentage distributions as appropriate. The number and proportion of events identified by: (a) the hospital codes compared with the case note review; (b) the discharge summary compared with the case note review; and (c) the hospital codes compared with the discharge summary, were calculated for each of the four geriatric syndromes and the eight medical diagnoses. In addition, variables were coded to identify whether or not a patient had any one of the geriatric syndromes, and any one of the medical diagnoses, according to the case note review, hospital codes, and discharge summary. Again, the number and proportion of events identified by the three sources of information were compared. Finally, tests of proportions were used to assess whether or not the discharge summaries and hospital codes identified patients with a geriatric syndrome as well as patients with a medical diagnosis.

Results

82 sets of notes were reviewed. There were 44 men and 38 women with a median age of 85.7 (range 76.4-102 years) and median length of stay of 19.5 days (range 4-118 days).

The case note review identified 59 patients with a geriatric syndrome and 62 with a medical diagnosis. Forty patients had at least one geriatric syndrome noted in the discharge summary; 16 (40%) were captured by coding. Of the 57 patients with at least one medical condition noted in the discharge summary, 52 (91.2%) were captured by the hospital coding (p<0.0001 for difference in hospital coding capture rates between geriatric syndromes and medical diagnoses) (Table 1).

Table I
Comparison of geriatric syndromes and medical diagnoses identified from case note review, discharge summary and hospital coding

The most common geriatric syndromes were falls (39 patients) and mobility impairment (45 patients). The most common medical diagnoses were pneumonia (27 patients) and urinary tract infection (27 patients). There was a marked difference between the capture of individual geriatric syndromes and medical diagnoses in both the discharge summary and the clinical coding. For example case note review identified 39 cases of falls. Only 16 (41%) of these were recorded in the discharge summary and, of these, only 3 (18.8%) were captured in the coding. In contrast, case note review identified 27 cases of pneumonia of which 23 (85.2%) were recorded in the discharge summary and 22 (91.7%) were captured in the coding (Table 1).

Of the geriatric syndromes, cognitive impairment was most likely to be identified in the case note review and discharge summary (77.4%) but of these cases, only 40% were also identified in the coding. Mobility impairment was the least likely to be identified in the case note review and discharge summary (11.1%) although pick up in coding of these cases was 80%. Of the medical conditions, there was 100% recording of stroke and peptic ulcer disease in the discharge summary but this completeness was only maintained in hospital coding for peptic ulcer disease. Atrial fibrillation and urinary tract infection were the medical diagnoses least likely to be recorded in the discharge summary (66.7% for both) but capture of this discharge summary information in coding was better (78.6%, 95% respectively) (Table 1).

Discussion

The management of geriatric syndromes forms a major component of modern clinical practice but our findings demonstrate poor capture of this information in both the discharge summary and hospital coding. The completeness of information for geriatric syndromes was significantly worse than for medical diagnoses and this suggests that the complexity of admissions for frail older people is not being reflected in coding data and that performance is likely to be underestimated.

The reasons are multiple14. There is evidence that the current coding system based on the International Classification of Diseases 10 (ICD-10) system is inadequate for classifying complex co-morbidity and geriatric syndromes15,16,17. The geriatric syndromes are generally ascribed ICD10 R-codes which are’ unattributable signs and symptoms’ and specific wording is required to describe the syndrome in order for a code to be ascribed. For example there is no ICD-10 code for intellectual deterioration. However confusion can be coded as unspecified dementia, delirium not induced by alcohol, disorientation not specified or as other dissociative disorders.

However the failure to record information on geriatric syndromes in the discharge summary suggests that education of hospital teams looking after frail older people is equally important. This may be a particular issue when discharge summaries are completed by non-specialist junior staff. There is evidence that the number of errors decrease with the seniority of the individual completing the summary17. Poor hand writing may also be contributory. One study showed that of 117 notes, 18 sets (15%) had handwriting so illegible that the whole clinical history was unclear18.

Several other studies have demonstrated systematic under-reporting of geriatric syndromes, for example one study showed that 38% of patient reported falls were not recorded on computerised accident and emergency records19 and it has been recently suggested that fall injury mortality is being underestimated20. This finding is consistent with previous work demonstrating that two common geriatric syndromes, incontinence and pressure ulcers, were rarely recorded on discharge and therefore could not be ascertained from hospital administration databases21. This is an area that is being actively addressed in the United States where the link between financial reimbursement and coding of hospital activity is now well established22.

There were some limitations to this small descriptive study. The number of cases for several medical diagnoses was low. Nevertheless it was possible to see a significant difference between the overall capture of information on geriatric syndromes and medical diagnoses. In addition, the retrospective case note review could not provide a definitive explanation why information in the medical notes did not reach the discharge summary or hospital coding.

However our findings do suggest approaches that could be used to improve the recording of information on geriatric syndromes. We are piloting a problem list bookmark which is kept at the current page in the medical notes and updated as new problems are identified. This is then used to complete the discharge summary. Also local education programmes are underway both on the importance of capturing the complexity of the care provided and on how this can be achieved. Limitations in the current coding system with regard to the consistent classification of geriatric syndromes remain to be resolved.

References

1. Walker S. Capturing health information -- a coding perspective. Bull World Health Org. 2006;84:254–5. [PubMed]
2. Lakhani A, Coles J, Eayres D, et al. Creative use of existing clinical and health outcomes data to assess NHS performance in England: Part 2--more challenging aspects of monitoring. BMJ. 2005;330:1486–92. [PMC free article] [PubMed]
3. Rowan K, Harrison D, Brady A, Black N. Hospitals’ star ratings and clinical outcomes: ecological study. BMJ. 2004;328:924–5. [PMC free article] [PubMed]
4. Early lessons from payment by results. London: Audit Commission; 2005.
5. Royal College of Physicians’ Information Laboratory The iLab project evaluation report. London: Royal College of Physicians; 2006.
6. Croft GP, Williams JG. The RCP Information Laboratory (iLab): breaking the cycle of poor data quality. Clin Med. 2005;5:47–9. [PubMed]
7. Smith SH, Kershaw C, Thomas IH, et al. PIS and DRGs: coding inaccuracies and their consequences for resource management. J Public Health Med. 1991;13:40–1. [PubMed]
8. Gardner M. Why clinical information standards matter. BMJ. 2003;326:1101–2. [PMC free article] [PubMed]
9. Rao P, Andrei A, Fried A, et al. Assessing quality and efficiency of discharge summaries. Am J Med.Qual. 2005;20:337–43. [PubMed]
10. Macaulay EM, Cooper GG, Engeset J, Naylor AR. Prospective audit of discharge summary errors. Br.J Surg. 1996;83:788–90. [PubMed]
11. Hasan M, Meara RJ, Bhowmick BK. The quality of diagnositc coding in cerebrovascular disease. Int J Qual Healthcare. 1995;7:407–10. [PubMed]
12. Dornan S, Murray FE, White G, et al. An audit of the accuracy of upper gastrointestinal diagnoses in Scottish Morbidity Record 1 data in Tayside. Health Bull.(Edinb.) 1995;53:274–9. [PubMed]
13. Campbell SE, Campbell MK, Grimshaw JM, Walker AE. A systematic review of discharge coding accuracy. J Pub Health Med. 2001;23:205–11. [PubMed]
14. Surjan G. Questions on validity of International Classification of Diseases-coded diagnoses. Int J Med.Inform. 1999;54:77–95. [PubMed]
15. Dixon J, Sanderson C, Elliott P, et al. Assessment of the reproducibility of clinical coding in routinely collected hospital activity data: a study in two hospitals. J Public Health Med. 1998;20:63–9. [PubMed]
16. Gray L, Martin FC. Classifying older patients in hospital. Age Ageing. 2005;34:422–4. [PubMed]
17. Macaulay EM, Cooper GG, Engeset J, Naylor AR. Prospective audit of discharge summary errors. Br.J Surg. 1996;83:788–90. [PubMed]
18. Rodriguez-Vera FJ, Marin Y, Sanchez A, et al. Illegible handwriting in medical records. J R.Soc.Med. 2002;95:545–6. [PMC free article] [PubMed]
19. Atri J, Pugh RN, Bowden D. Are we falling at the first hurdle? Estimating under-recording of falls in Accident and Emergency. J Public Health (Oxf) 2005;27:33–5. [PubMed]
20. Jansson B. Coding errors and underestimation of fall injury mortality. Am J Public Health. 2005;95:1305–6. [PubMed]
21. Berlowitz DR, Brand HK, Perkins C. Geriatric syndromes as outcome measures of hospital care: can administrative data be used? J Am Geriatr.Soc. 1999;47:692–6. [PubMed]
22. Chattopadhyay A, Bindman AB. Accuracy of Medicaid payer coding in hospital patient discharge data: implications for Medicaid policy evaluation. Med.Care. 2005;43:586–91. [PubMed]