TableA.5 Summary of articles examined for economic burden of GI disease

ID and AuthorsResearch setting and year of studyStudy designSample sizeTopic of documentKey results and conclusionsComments
Bassi et al708UK (2004)A six month cohort study307 Cases of ulcerative or undetermined colitis and 172 cases of Crohn’s diseaseProfile, determinants and scale of cost of IBD in UKInpatient services were required by 14% of the sample but accounted for 49% of secondary care costs. Drug costs accounted for less than a quarter of total costs. Individual patient’s costs ranged from £73 to £33 254.Mean cost per patient were £1256 (CI 988 to 1721) for colitis and £1652 (CI 1221 to 2239) for Crohn’s disease. The corresponding average cost for the ambulatory was respectively £516 (CI 452 to 618) for colitis and £539 (CI 497 to 589) for Crohn’s disease. For the hospitalised group the mean costs were £6923 (CI 5415 to 8919) for colitis and £7658 (CI 5693 to 10561) for Crohn’s disease. The mean cost of an “incident” was respectively, £2662 (CI 1006 to 5866) for colitis and £2111 (CI 1488 to 3078) for Crohn’s disease.Survey suggested that the average costs for six months were <£30 per patient for primary care visits, and the median loss of earnings was £239 for colitis and £299 for Crohn’s disease.The mean (range) for out-of-pocket expenses, prescription charges, OTC drugs, etc, was £40 (0–250) and £66 (0–750) for colitis and Crohn’s diseaseLevel and profile of healthcare costs for IBD in Britain is remarkably limited. Robust CEAs data for rival treatments are lacking as few studies incorporated the prospective collection of resource data. Furthermore, the cost inputs for modelling exercises have relied on subjective cost estimates rather than real patient data. Diagnostic information is not routinely collected for most ambulatory care episodes.Also, there is no requirement to record patient borne costs.
Wells et al739UK (1995)ReviewNAThe burden of IBSThe study used the prevalence approach to estimate the NHS resource use. The number of GP visits was estimated on the Office of Population Census and Surveys, study of morbidity in general practice in England and Wales (1995) and indicated that 846 349 visits a year in the UK are IBS related. This represents about 10% of the total primary care workload for GI diseases and results in an expenditure of £13.1m per year. Prescriptions for all medicines for IBS were estimated using the DIN-LINK database; a total number of 1.7 million prescriptions were written for IBS in 1995. This figure multiplied by the average net ingredient cost for each prescription for gastrointestinal medicines (£5.1) and including an average of £1 per prescription yields a total expenditure on GP-prescribed medication for IBS of £10.5m. This figure was then adjusted to £12.5m to account for the difference between the prevalence estimated from the database used and the general practice survey.Cost in secondary care (outpatient and inpatient stays) were estimated at £16.4m. In addition it was estimated that 3600 people are admitted to hospital, generating a total of 18 000 beds days for a total of inpatient costs of £3.4m. Hence, the total cost to the NHS of 45.4 million a yearThe authors give a warning about the accuracy of the estimates—the prevalence approach method relies on ICD coding. In those diseases where an unambiguous definition of the condition exists and which have a unique identification number in the ICD, it is a straightforward process. This is not the case for IBS, although one ICD number does refer to irritable colon, it is uncertain how accurately the volume of resource utilisation coded under this heading reflects the true magnitude of IBS
Akehurst et al740UK (6 GP surgeries in the Trent Region) (1999)A case-control study374Impact of irritable bowel syndrome on time off from work, utilisation and cost of health services and HRQoLPatients with IBS had considerably lower HRQoL than controls. They scored worse in all dimensions of the SF-36 and the EQ-5D. The IBS group had 0.2 (95% CI −2.1 to 2.6) more days off than the control group; this difference did not reach statistical significance. A significantly higher number in the control group had no time off from work in the previous three months, and more people in the IBS group had more than a week off in the previous three months. On average, the patients with IBS cost the NHS £123 (95% CI 35 to 221) more each year than people in the control group (p = 0.04). The cost figures show that even using a conservative estimate of 9% of IBS prevalence, the total cost for the UK exceeds £200m a year, and for a 25% prevalence rate would rise to £600m a yearIt is estimated that only about 30% of these are referred on to specialists. Further research would be needed to determine the impact of IBS on the quality of life of those affected who do not present to their GP
Creed et al741UK (northern England) (1998)Cross sectional survey257 Patients (from secondary and tertiary GI clinics) who did not respond to the usual treatment and were recruited for a trial of psychological treatmentTo determine whether the severity of bowel symptoms and psychological symptoms directly influence HRQoL and healthcare costsGlobal severity and somatisation contributed to the physical component score of the SF-36, but only psychological scores were associated with disability due to ill health. These variables did not predict healthcare costs (R2 = 9.3%). Resource use costs were collected prospectively; these were $1743±2263 and included$1338±2107 for secondary care costs, $410±424 for primary care costs,$2.7±30.8 for alternative treatments; $63.5±260 in patient borne costs; and$334±1052 from loss of productivityThis study is only representative of patients with IBS whose management is difficult because of refractory symptoms, marked disability, and high care utilisation.Further work is needed to assess what leads to high healthcare costs in these patients
Lewison483UK (1997)Use of existing data from ONS (and counterparts in Scotland, NI) and from the DSS on work related dataNABurden of GI disease studyCost of lost productivity due to early death from GI disease  =  £2.2 billion, from long term disability  =  £0.8 billion, from short term disability  =  £2.0 billion, NHS costs  =  £3.0 billion (45% in patient cost, 27% drugs). Total burden of disease  =  £8 billion (1997)Crude estimate. There is a great debate on how sickness absence should be costed (human capital v friction cost method) as these methods give very different figures. NHS costs, especially drug costs, are here more reliable.Much debate on value of burden of illness studies
Keighley324UK (2003)ReviewGI cancers in EuropeCrude statements about what the main cost components are for the treatment of a variety of GI cancers. Most detail given for colorectal cancer. Average cost of treating a case of colorectal cancer in the UK estimated at €12 630Crude estimates. Interesting for intercountry comparisons, but very inaccurate
Belsey516UK (2001–02)Modelling exercise based on Hospital Episode Statistics of England.109 000Prescribing practiceOf a population of 109 000 patients in the UK in 2001–02 waiting for knee and hip replacement, around 637 had upper GI bleeding linked to NSAID treatment, with between 51 and 89 deaths resulting. Based on the cost of £3000 (estimated by Moore et al, 1999) for every admission for an NSAID related GI bleed. The additional 637 events identified in this analysis will therefore account for an expenditure of £1.9mA critical review of prescribing practice is required, in order to reduce reliance on agents with a tendency to trigger upper GI ulceration
Moore and Phillips709UK (1999)Modelling exerciseNAThe burden of NSAID related gastrointestinal diseaseThe burden of NSAID adverse effects to the NHS was calculated for the UK population on the basis of an average PCG with a population of 100 000. The average cost per year was estimated as: (a) low cost mix  =  £241 (10% coprescriptions); (b) middle cost mix  =  £266 (15% coprescriptions), and high cost mix  = £308 (20% coprescriptions). The UK inpatient costs related to UGI bleeding were estimated at £35m (£7.7 per patient). Because aspirin and ibuprofen are OTC drugs, the total burden is likely to be greaterMost of the information on presription rate was collected from the USA.There is a high variation of prescribing mix in the UK, which we could not account for in detail. Given the high cost of coprescribing, a systematic survey is needed urgently
Duggan743UK (1998)Modelling exerciseNACost of management of UGI diseaseThe cost per patient (prescription costs, GP surgery visits, and outpatient services) were as follows: IGPCG  =  £157, Hp testing for all  =  £175, Hp testing and endoscopy for all  =  £236, Hp testing for ulcer  =  £172, Hp testing and endoscopy for ulcer  =  £219, endoscopy for all  =  £404, endoscopy for patients aged over 45  =  £335, and endoscopy for patients aged over 45 presenting for the first time  =  £218.All the guidelines modelled here would result in a significant increase in the number of patients presenting for endoscopy. The scenario involving endoscopy for all patients aged over 45 would require a 13-fold increase in the provision of endoscopy services. A one year period was chosen because of the lack of studies with a longer follow-up. Hospitalisation costs could not be included because of insufficient published data to quantify any differences between differing drug treatment regimens. Because the IGPCG did not include gastric cancer, this was also excluded in the five scenariosBearing in mind the limitations of any modelling exercise, this study used robust data sources
Morant et al744UK (Scotland) (1989–95)Population based observational cohort study17 244 New users of aspirin, each with 10 matched comparatorsTo determine the cost to the NHS of prescribed low dose aspirinAspirin use cost an additional £49.86 a year, made up of £1.96 for aspirin tablet (4%), £5.49 for dispensing costs (11%), £24.60 for UGI complications (49%), and £17.81 for renal complications (36%). These results projected to Scotland would give an estimate of £3 m (fastidious analysis) or £11 m (pragmatic analysis) a year for newly treated patients.If we assume that the antiplatelet trial meta-analysis is an accurate assessment of the benefits of aspirin, then the cost for preventing one vascular event lies between £62 500 (primary prevention) and £867 (secondary prevention—patients without history of UGI or renal problems)A large number of sensitivity analyses were performed and they all confirmed the original results
Haycox et al697UK (1996)Modelling exerciseNAThe resource demand placed on general practitioners by patients following the many different management strategies available for UGI diseaseThe average cost incurred in primary care by patients with symptoms of UGI is £165; however, there was a wide variation across patients (range 10 to 989). The IMS database indicated that an average of 4.4 prescriptions a year is currently provided to patients presenting to GPs with UGI complaints. Implementing the IGPCG algorithm it reduces the number of prescriptions to an average of 1.9, with 70% of patients treated in accordance with the algorithm receiving long term (more than six months) symptom relief from their course of treatment. This would save £70 m out of current total drug cost of £488 m. Such cost reductions arise mainly from improving diagnosis and symptomatic management of patients and the more intensive eradication of H pylori infectionAlthough the results were confirmed by an extensive number of sensitivity analyses, the authors consider the study a preliminary analysis because refinement of the model is still continuing
Haycox et al742Europe and USA (1996)Modelling exercise4 CountriesThe extent to which economic analyses can be transferred across national bordersAssuming UK  =  100 (£174), the cost of managing UGI for some of the other countries was respectively: Sweden  =  179(£311), Germany  =  101(£176), Switzerland  =  163(£284)In all major cost categories the cost of treating a patient in the UK and Germany is significantly lower than treating the patient in Sweden and SwitzerlandThe fact that the pound sterling was comparatively strong compared with the other currencies might have contributed to the lower comparative costs identified in the UK.Further research is required to identify in more detail the factors influencing such variations and their impact both upon patients and public expenditure
Thomson and Booth653UK (1996)ModellingNAThe resource implication of TD (traveller’s diarrhoea)The incidence approach is used, the incidence of TD is assumed to be 8, 15, and 56 in low, intermediate, and high risk countries, respectively. The cost varied from £78m (prophylaxis) to £17m (treatment)These calculations are rough and reflect the poor state of knowledge in this area
Ellis768UK (1989/1990)Observational study3975 Usable dataAssess the contribution of the most frequently performed procedures to surgical workload and evaluate the financial implicationsThe patients incurring the greatest costs were those who had undergone large bowel surgery, vascular reconstruction, or amputation. The top five procedures in order of frequency were upper GI endoscopy, inguinal hernia repair, cystoscopy, tranurethral resection of the prostate, and surgery for long saphenous varicosities.Upper GI endoscopy in 420 patients (352 as day cases) cost £149 630. The cost of hospital stay for patients with large bowel surgery for carcinoma was £38 529 for 12 patientsThe accuracy of coded discharge data on inpatient care is still a problem; 18% of the records had to be excluded because they were imprecise, ambiguous, or not present at all.The article fails to compare its results with others and fails to assess the external validity of the findings
Beard et al769UK (Costing data) (1990–98)Literature review and modelling exerciseNACost of liver resectionBased on the average costs from the Royal Hallamshire Hospital, Sheffield, the total treatment costs with resection were £6402. The cost of the alternative pathway of care (chemotherapy, Gramont regimen) is estimated at £6669. Hence the marginal cost saving of resection is £267. However, if patients have recurrence, the resection would only delay the chemotherapy by one year. Hence it would be cost saving only with no recurrence (patient remained free of tumour).Excluding any savings that may be made through avoided chemotherapy, and assuming no differences in salvage treatment of relapses, the cost per LYG ranges from £2134 to £3945. Even if 50% of the patients were found unsuited for surgery, or only received palliative resection, the survival benefits for the remaining patients are likely to cost around £6078 per LYG (undiscounted five year survival)The study results might be highly sensitive to Sheffield unit characteristics. Moreover, some of the data are based on anecdotal evidence.There are no RCTs which directly compare the role of liver resection against treatments for liver metastases
Sheridan et al745UK (1991)Prospective cohort study.100 Patients aged between 15 and 35 admitted with lower abdominal pain to one general surgical firmTo audit the extent of the problem of NSAP (non-specific abdominal pain).To assess resource implications67 Patients were disgnosed as having NSAP; GI disorders were one of them. The total cost to the NHS of these patients was £54 115.On the basis of this, it is calculated that NSAP might be responsible for a total of 7708 emergency general surgical admissions per year in Wales, using 31 757 bed days and costing the NHS in Wales 6.4 million. Extrapolating these figures to the UK as a whole, the annual cost of NSAP to the NHS may be over £100m a yearThe study provides a crude estimate of the NHS costs, in particular the extrapolation from one DGH to the country as a whole
McLoughlin et al746Europe (2002)Literature review10 CountriesThe burden of coeliac diseaseIn England it was estimated that 120 000 patients are affected by the condition and the test and treatment costs were £120m and £216m, respectivelyThis is a short article to provide a “snapshot” of coelic disease. No formal assessment of the robustness of the data sources is offered
Mahmood and McNamara16UK (Ireland) (2003)Modelling exercise. Using the Swedish model of cost analysis recently published in Pharmaeconomics 2002, the authors applied the same model to calculate the burden of the disease in different countries across Europe6 CountriesBurden of GERD (gastro-oesophageal reflux disease)The UK prevalence was estimated as 24% giving the following costs: direct costs (million euro)  =  2591; indirect costs  =  1610; drugs  =  1532; cost of sick leave  =  1358; and other costs  =  1239It is difficult to calculate the economic burden of GERD or peptic ulcer disease as most studies have used the broader term “dyspepsia” in their calculations.The study uses the prevalence approach, and there is insufficient information on the data sources for costing and how the summary figures are arrived at
Moayyedi and Mason747UK (2001)Article review8473 Individuals participating in the HELP study.To identify NHS costs, 5056 primary care notes were reviewed for 1992–94Clinical and economic impact of dyspepsiaDyspepsia was costing £21 per person per year—hence about £1000m each year in the UK. Within primary care dyspepsia was costing £11.25 per person per year, which represents £500m each year in the UK.Although based on patients in the Leeds HELP study, the study is very robust. The authors do point out that the national figure is representative as long as the cohort is representative of all the populationCost effectiveness management strategies and treatments are urgently required. Strategies for managing dyspepsia have focused on attempting to reduce the endocscopy workload, although this procedure accounts for only a small proportion of the total costs of dyspepsia. Future approaches should also examine the drug used, as this might result in more cost savings
Logan and Delaney748UK (1999)Review articleNAImplications of dyspepsiaAt any one time 4% of the population are thought to be taking drugs prescribed for dyspepsia. Drugs for dyspepsia account for 10% of drug expenditure. The number of gastroscopies performed each year is also rapidly increasing; 450 000 tests were performed in the United Kingdom in 1996This is a short article and the author does not provide any information on the data sources
Somasekar et al749UK (Wales) (1999–2000)Retrospective study and model exercise156 Patients who underwent elective cholecystectomy from a DGHThe economic burden of patients waiting for cholecystectomy admitted with recurrent gallstone related symptomsThe mean (SD) waiting time for surgery was 12 (3) months. 37 patients were admitted as an emergency owing to gallstone related symptoms and complications while waiting. The cost for each episode was £946 and the total cost of treating the 37 patients was calculated to be £44 462. Performing early laparoscopic cholecystectomy for acute cholecystitis may help to reduce costs by preventing recurrent emergency admissions of these patientsCrude estimate. A small analysis of the effect of comorbidities is also made