Percutaneous endoscopic gastrostomy (PEG): The unseen mortality

Percutaneous endoscopic gastrostomy (PEG): The unseen mortality


303KB Sizes 0 Downloads 19 Views

G A S T R O E N T E R O L O G Y Vol. 114, No. 4


• G0154 MEDICAL OR SURGICAL THERAPY FOR EROSIVE REFLUX ESOPHAGITIS: A COST-UTILITY ANALYSIS. DC Sadowski, J. Romagnuolo, M. Meier, Departments of Gastroenterology and General Surgery, University of Alberta, Edmonton, Alberta, Canada. BACKGROUND AND AIMS: Medical therapy with proton pump inhibitors for endoscopically proven reflux esophagitis has been widely studied and is a safe and effective treatment option. Of late, the surgical treatment of choice lbr this disease has become the laparoscopic Nissen fundoplication (LNF). The aim of this analysis was to compare the cost and utility of healing and maintenance regimens of omeprazole with laparoscopic Nissen fundoplication in the framework of the Canadian medical system. METHODS: A cost-utility analysis was performed comparing the two abovementioned strategies. A two-stage Markov model was used to obtain cost and efficacy estimates in a healing and maintenance phase with a time horizon of five years. Our base case was a 45 year-old male with erosive reflux esophagitis refractory to H2-blockers. Discounted direct costs were estimated from the perspective of a Canadian provincial health ministry. Direct costs of LNF were calculated using physician fees, consumables, equipment costs and length of stay data. Quality of life estimates were derived from the medical literature. Sensitivity analyses were performed to test the validity of assumptions and determine thresholds. RESULTS: LNF was less expensive than omeprazole for the five year period studied ($3225 vs. $5695 per patient). The LNF strategy became the most cost-effective option at 2.4 years of follow-up. One-way sensitivity analysis on drug costs found that omeprazole 20 mg per day would have to cost less than $35 per month (approximately half its current cost) before long term maintenance therapy becomes cost-effective. The cost of LNF would have to be more than $5561 before medical therapy became more cost-effective. Estimates of quality-adjusted life years were not significantly different between the two groups. CONCLUSION: For patients with severe esophagitis, LNF is a cost-effective treatment option to long term maintenance therapy with proton pump inhibitors. •


"REASSURANCE COST" MAY AFFECT ESTIMATES OF RESOURCE UTILIZATION IN DYSPEPSIA IF GASTROSCOPY IS INSUFFICIENT TO PREVENT FURTHER TESTING. AV Sahal. I Penman, D Williams, G Mishra, W Knapple, A vanVelse, RH Hawes, and BJ Hoffman. Digestive Disease Center, Medical University of South Carolina, Charleston, SC. BACKGROUND: Cost-analyses of dyspepsia management do not account for additional resource-utilization that may occur when a negative gastroscopy[EGD] is insufficient to reassure patients and/or physicians that no serious disease is present. AIM: 1) What is the frequency of negative and/or non-contributory abdominal imaging tests in dyspepsia patients (pts)? 2) Do EGD pts undergo less additional testing than nonEGD pts? 2) What is the cost of reassurance? METHODS: All dyspeptic patients seen for endoscopic ultrasound completed a structured physician-administered questionnaire on previous dyspepsia-related negative and/or non-contributory testing. Facility direct and indirect costs for barium meal [UGI] $165, abdominal ultrasound [US] $141, abdominal CT [CT] $238, and abdominal MRI [MRI] $376 were used to calculate the "reassurance cost"/patient. RESULTS: 121 pts studied since 05/01/97 (study ongoing). Dyspepsia types: pancreatic 55%; non-specific 14%; biliary 13%; IBS 8%; ulcer 6%; reflux 4%. Ninety-three (76%) had EGD; 29(34%) did not. "EGD" and "noEGD" gps were not significantly different for age, sex, race, and dyspepsia type; but differed in mean 2D (18+/-14mo vs 9+/-14mo, p=0.003). However, the # tests (EGD or nonEGD pts) did not correlate with ED. Total # tests/pt (incl. EGD): 5 tests: 3%; 4: 23%; 3: 33%; 2: 25%; 1: 10%; 0: 7%. Frequency of tests in EGD and non-EGD pts: Previous UGI

Previous US Previous CT Previous MRI % tests (exelud. E G D )

ALL (n=121) 33.1% 65.2% 80.2% 9.1%

EGD (76%) 38.0% 68.5% 84.8% 8.9%

NoEGD(34%) 17.2% 55.2% 65.5% 10.3%

1.9 + l. 1

2.0 =e 1.0

1.5 -+ 1.l


PREVALENCE AND PROFILE OF TUBERCULOSIS IN CHRONIC LIVER DISEASE PATIENTS. S Salgal, HP Nandeesh, SR Agarwal, A Misra, S K Jain, SK Satin. Dept. of Gastroenterology, G.B. Pant Hospital and New Delhi TB Centre, India.


Obiectives: Tuberculosis (TB) poses difficult problems in patients with chronic liver disease (CLD), both in diagnosis - due to its varied overlapping symptoms, and, in management - because of the high propensity of hepatotoxicity. The prevalence of TB in CLD patients therefore remains unclear. A prospective study was carried out to determine the prevalence and clinical profile of TB in CLD patients. Methods: Consecutive CLD patients seen between June, 1995 to Nov, 1997 were screened for TB based on the following parameters: fever, cough for > 2 wks, hemoptysis, unexplained weight loss, increasing ascites not responding to diuretics, unexplained bowel symptoms, radiological lesions suggestive of TB, and past or family history of TB. Investigations included Hb, ESR, mantoux test, sputum AFB, cytology, biopsy, chest X-ray, US abdomen, CT scan and barium. Adenosine deaminase ( > 32.2 U/L) and PCR for M. tuberculosis in ascitic fluid were also assessed. TB was diagnosed based on one or more of the following: i) histological evidence of caseating granuloma or AFB ii) sputum positive for AFB, iii) growth of M. tuberculosis on culture, iv) good response to chemotherapy. Ascitic TB was diagnosed based on raised counts with lymphocyte predominance, ]'protein ( > 2.5g%), ]'ADA, with or without PCR positive for M. tuberculosis. Prevalence of TB in the general population from the registry was 1.6%. Results: A total of 736 patients of CLD were seen of which 60 (8.1%) were diagnosed to have one or the other form of TB. The mean age of the patients with TB was 38 years with a male predominance (M:F=4:I). Fever was a common presenting symptom (60%). Only 1/3rd presented with a history of fever > 6 mo. Anorexia (65%), weight loss (43%) and cough (45%) were also common. Past history of TB was present in 5 (8.0%) patients. Abnormal pulmonary findings were clinically seen in half the patients. The common presentation of TB included: pulmonary (55%), abdominal (26%), disseminated (12%), lymph nodal (5%) and genitourinary (2%). Majority of the patients were in Child's grade B or C (27 and 19 respectively). Alcohol (38%) and hepatitis B (30%) were the important etiology for CLD. Abdominal TB was more common in Child's A patients while pulmonary TB was more common in Child's B or C as compared to Child's A (p<0.05). 32% of abdominal TB patients had associated pulmonary TB while only 18% of pulmonary TB patients had vice versa. Conclusions: (i) Our results suggest that there is nearly five times higher prevalence of TB in CLD patients (8.1%) compared to the general population (1.6%), pulmonary TB being the commonest form, (ii) profile of TB is often similar to presentations of the liver disease p e r se, hence a high index of suspicion and good screening methods are needed to identify TB in patients with CLD. • G0157 PERCIYrANEOUS ENDOSCOPIC GASTROSTOMY (PEG): TIlE UNSEEN MORTALITY. D.S Sanders 1, M.J. Carter 2, J.D'Silva3, G. James t, R. P. Bolton 1, K.D. Bardhan 3. Departments of Gastroenterology at IDoncaster Royal Infirmary, 2The Royal Hallamshire Hospital, Sheffield and 3Rotherham District General Hospital; UK. Background and Aim: PEG has revolutionized long-term enteral feeding and has improved survival in acute dysphagic stroke and oropharyngeal malignancy. However, its simplicity has lead to its use in areas of uncertain benefit. We therefore assessed survival in our patients with particular reference to the indication for PEG. Patients and Method: The cohort is of 361PEG patients treated between August 1992 - July 1997 at 2 district general hospitals and comprised 4 subgroups as follows. n

P 0.04 NS 0.03 NS 0.02

Mean "reassurance cost" per pt: EGD pts: $394; nonEGD: $298. CONCLUSIONS: 1) Dyspepsia pts undergo multiple negative and/or noncontributory abdominal imaging tests -- with no relation to symptom duration. 2) EGD pts actually had more tests than nonEGD pts; suggesting that EGD is insufficiently reassuring to prevent further testing. 3) On a large scale, "reassurance cost" may be substantial (since dyspepsia is a common problem). It should therefore be considered in economic analyses of dyspepsia management. 4) In dyspepsia pts who are referred for EGD, an imaging test that can comprehensively examine the upper GI tract and surrounding structures in one sitting (e.g. endoscopic ultrasound) may reduce overall costs and resource utilization if its cost is sufficiently low and it prevents further testing. Funded in part by ASGE-Olympus Advanced Endoscopic Training Scholarship & by ADHF Outcomes Training

Group 1


Mean Age (Y) 63.7

Group 2



Group 3



Group 4



PEG Indication Ompharyngeal malignancy: PEG before surgery Acute dysphagic stroke: CT proven or clear clinical signs Anorexia with cognitive impairment: Dementia: Unable/unwillingto eat Miscellaneous: Head injury, Multiple sclerosis, Motor neurone disease

Survival was calculated by life table analysis (Kaplan-Meier). Results:

Group 1

2 3 4 Overall

Probability of survival (%) at 1 Mo 3 Mo 6 Mo 86 72 64 78 64 54 46 22 19 86 76 65 72 56 48

12 Mo 50 44 10 52 37

2 Yrs 37 31 7 39 27

4 Yrs 22 22 0 25 16

April 1998 *

Clinical Practice A39

Group 3 had a particularly poor prognosis compared with other patients (P < 0.0001, Mantel-Cox Logrank test); 1/2 had died in 1 month and nearly 9/10 at 1 year. In the other three groups, approximately 1/3 were alive at 2 years, and most of them survived beyond this.

Conclusion: 1) PEG had high initial mortality in all groups, consistent with previous studies. 2) This is the first demonstration that patients with anorexia and cognitive impairment (almost 1/3 of our patients) do particularly badly. Consequently, in them we now advise against PEG feeding. • G0158

A SIX-WEEK RANDOMIZED CONTROLLED TRIAL OF OLESTRA OR REGULAR SNACK FOODS: COMPARISON OF GASTROINTESTINAL SYMPTOMS IN A LARGE POPULATION. R.S. Sandier, University of North Carolina, Chapel Hill, NC; N.L. Zorich, T.G. Filloon, H.B. Wiseman, D.J. Lietz, M.H. Brock, R.K. Miday, Procter & Gamble, Cincinnati, OH. Background: Olestra is a new fat replacement that adds no calories to food.

Because olestra is neither digested nor absorbed it has the potential for GI symptoms. We conducted a randomized, double-blind trial to determine the true incidence and impact of GI symptoms in free-living adults and children consuming olestra snacks ad libitum. Methods: 3,181 participants were randomized by household and consumed either olestra or regular full fat potato and corn chips in olestra-labeled packages. Participants could also choose additional full fat snacks in commercial packages. A household representative visited one of two central sites weekly, for 6 weeks, to select snacks and to return daily diaries. Each subject in the study recorded daily the amount of snacks consumed, GI symptoms, and the impact of symptoms on activities. Results: A total of 2,296 teens and adults (ages 13-89) and 885 children (ages 2-12) consumed test snacks during the study. There was no overall difference in GI symptoms. Of 1,620 individuals in the olestra group, 619 (38.2%) reported one or more GI symptoms, compared to 576 (36.9%) of the 1,561 individuals consuming full fat snacks, p=0.60. There were also no significant differences between the olestra and full fat groups in the incidence of any one of 8 GI symptoms, except for increased nausea in the full fat group (Table). Mean number of symptom-days were not different between groups for any of the 8 GI symptoms, except for "more frequent bowel movements" (3.7 vs 2.8 days, olestra vs. full fat, p=0.04). There was no difference in the impact of symptoms on daily activities between the 2 groups. Of subjects who reported that they visited a physician for GI symptoms, 6 were in the olestra group and 10 in the full fat group. Test chips were consumed frequently throughout the study (mean 20.2 vs. 21.6 of 42 study days, for olestra and full fat). The frequency of consumption did not correlate with symptom reporting in either group. Symptom

Heartburn Nausea Vomiting Gas Bloating Abdominal cramping/pain More frequent BM Looser stool

Olestra n = 1,620

Regular n = 1,561


8.6% 5.7% 1.8% 24.2% 11.2% 15.0% 20.5% 25.3%

8.4% 8.4% 1.8% 21.7% 9.4% 15.1% 17.4% 23.1%

.88 .02 .99 .25 .17 .94 .11 .31

Conclusions: Ad libitum consumption of olestra corn and potato chips over a

six week period is not associated with an increase in incidence or severity of GI symptoms. Authors are a research consultant (RS) or employees of Procter & Gamble • G0159 DIGESTIVE COMPLAINTS IN THE UNITED STATES: A NATIONAL SURVEY. R.S. Sandier, Center GI Biology & Disease, Univ. of North Carolina, Chapel Hill, NC; W.F. Stewart, J.N. Liberman, J.A. Ricci, Innovative Medical Research, Tnwson, MD; N.L. Zorich, Procter & Gamble, Cincinnati, OH. Background. Although episodic digestive complaints, including lower abdominal pain, bloating and loose stools are thought to be common, there are no published estimates of their prevalence in the US population. Valid estimates would provide useful information about the public health impact of digestive complaints and a context for interpreting reports of food-born illness or symptoms following the ingestion of certain foods and drugs. Methods. The study was a national cross-sectional telephone survey of adults. The response rate among eligible households was 70.7%. Respondents were questioned about lower abdominal pain or discomfort, bloating or distention, and loose stools or diarrhea in the month prior to interview. Affirmative responses to any one symptom were followed by questions about the frequency, duration, severity and impact of symptoms, and use of medical care, physician visits and medications. Reliability of symptom reporting was assessed by re-interviewing a random sample of 155 subjects with an average 13-day interval between interviews.

Results. Among 2,510 respondents, 1,017 (40.5%) reported one or more digestive symptoms within the month prior to the interview. Abdominal pain was reported by 21.8%, bloating by 15.9% and loose stools by 26.9%. Women were more likely than men to report abdominal pain (24.4% vs. 17.5%) and bloating (19.2% vs. 10.5%) but not loose stools. Adjusted prevalence ratios comparing women to men for abdominal pain, bloating and loose stools were 1.5 (95% CI 1.1-1.9), 1.9 (1.5-2.4), and 1.0 (0.9-1.2), respectively. Symptoms were less common among those age 60+. Among those with symptoms, more than 70% rated them as moderate or severe in intensity and more than 20% reported significant limitations to daily activities. More than 20% reported a GI symptom in the past 24 hours; 14% consulted a physician because of their symptom(s) and nearly half treated their symptom(s). The reliability study showed 79 to 90% agreement between interviews in occurrence of symptoms and related characteristics. In applying age- and sex-specific monthly prevalence rates from the survey to the US population, we estimate that approximately 79 million individuals experience abdominal pain or bloating or loose stools at least once a month; the numbers who experience moderate to severe abdominal pain, bloating, and loose stools at least once a month are 30 million, 22 million, and 49 million, respectively. Conclusions. More than 40% of the US adult population experience one or more digestive complaints each month. The symptoms are often perceived as moderate or severe in intensity and often significantly limit daily activities. This research was funded by Procter & Gamble, Cincinnati, OH. G0160

I N C I D E N C E OF GASTRIC AFFECTIONS IN A COUNTRY-LIKE PART OF GERMANY. M. Schumacher, K. Fiinfhausen, V. Hug, H. Wetzel, study group Wolmirstedt, Germany Introduction: Knowledge about the incidence of disease is a major parameter

for the assessment of treatment success and prospective working. On condition of successful strategies for eradication of H. pylori a decrease in incidence of gastric and duodenal ulcer is expected. In our region a specialist for gastroenterology and 26 general practitioners are working closely together so that 90% of the diseases in this region are diagnosed by that group of physicians and data of about 30.000 patients are recorded. Aim: The aim of this population-based study was to determine the incidence of gastric affections among patients in primary health care. Method: Patients complaints and endoscopic findings were differentiated in dysmotility-like dyspepsia (main symptom: acid reguritation, flatulence, early satiation), ulcer-like dyspepsia (main symptoms: pain relieve by food), refluxlike dyspepsia (main symptom: heartbum), gastric and duodenal ulcers, esophagitis grade I-IV (by Savary Miller), gastric carcinoma. Their incidences were calculated and a monthly evaluation was done for dyspepsia, esophagitis, gastric and duodenal ulcers. The total number of consultations was recorded in relationship to age and sex and the proportion of gastric affections was calculated. Results: Table 1 shows the figures for the incidences of different forms of esophageal and gastric affections: incidences in 1996 (per 10,000 inhabitants) esophagitis grade I-IV dysmotility-like dyspepsia ulcer-like dyspepsia reflux-like dyspepsia gastric ulcer duodenal ulcer gastric carcinoma

22 244 64 32 10 28 1

In men between 51-60 years of age and in women between 41250 years of age were at a higher risk of developing duodenal ulcer. In patients > 40 years of age esophagitis and gastric ulcer were significantly more frequently diagnosed. Reflux-like dyspepsia occured to an equal extent in men and women. Women > 40 years of age were at higher risk to develop dysmotilitylike dyspepsia. 98% of all duodenal ulcers were H. pylori postive. One percent of all H. pylori positive patients fell ill with a severe gastric disease like duodenal and gastric ulcers and carcinoma upon an estimated prevalence for H. pylori infection of 30%. Conclusion: In this part of Germany an eradication rate of 32% failed to lower the incidence of 28-30 duodenal ulcers per 10,000 inhabitants in the period of 1995-1997. Therefor it is recommended that H. pylori-eradication should be performed frequently in patients with ulcer-like dyspepsia.