Percutaneous endoscopic gastrostomy (PEG): change in practice since 1988

BACKGROUND AND AIMS

We previously reported a 30-day mortality following percutaneous endoscopic gastrostomy (PEG) of 8% (1988-92). Concerns over increasing mortality rates prompted us to survey current practice compared with 1988-92: assess case mix, outcome, risk factors for early death, and review practice guidelines.

METHODS

78 consecutive adults were referred for PEG over 7 months. Baseline characteristics, including age and functional status (Barthel Index), and outcome at 30 and 180 days were prospectively evaluated.

RESULTS

74 patients. Median age 69 years; male 55%. Major underlying diagnoses: cerebrovascular disease 42%, head and neck tumours 19%, motor neurone disease 4% (33%, 16% and 27% in 1988-92). Mortality rates at 30, 90 and 180 days were 19%, 35% and 42% respectively (8%, 20% and 37% in 1988-92). Univariate analysis showed that age >75 years, Barthel Index <1 and Glasgow Coma Scale < or =10 were significant risk factors for death at 30 days: odds ratios (95% confidence intervals) 3.9 (1.1-13), 5.9 (1.4-25) and 4.4 (1.2-15) respectively.

CONCLUSIONS

30-day mortality was increased from 8% to 19% between 1988-92 and 1998-99 reflecting a change in referral patterns: more elderly with cerebrovascular disease and fewer with motor neurone disease. Age and functional status should be considered when advising on PEG feeding.

The diagnosis of small bowel bacterial overgrowth in elderly patients

OBJECTIVE

To examine the frequency of small bowel bacterial overgrowth in elderly subjects.

DESIGN

Prospective observational study.

SETTING

Department of Health Care of the Elderly in a university teaching hospital.

PATIENTS

Thirteen elderly control subjects having undergone normal gastroscopy with normal hematology and anthropometry; 39 consecutive referrals of elderly patients with clinical or biochemical evidence of malnutrition.

MAIN OUTCOME MEASURES

Duodenal bacterial counts and the 14C-glycocholate, hydrogen, and 14C-xylose breath tests.

RESULTS

Five of the 13 control subjects and 21 of 39 patients with malnutrition had high duodenal bacterial counts (greater than 10(5) organisms/mL), whereas only seven patients were judged to have clinically significant bacterial overgrowth. The 14C-xylose breath test showed a high specificity (89%) but a low sensitivity (30%) when compared with the results of duodenal culture. The hydrogen breath tests and the 14C-glycocholate test showed a similar picture of moderately high specificity and low sensitivity (77% and 20%, respectively, for the hydrogen breath test, 76% and 33%, respectively, for the 14C-glycocholate test).

CONCLUSIONS

A positive xylose breath test was the best predictor of high duodenal bacterial counts. However, since many elderly subjects have high duodenal bacterial counts, the overdiagnosis of bacterial overgrowth syndrome will result if duodenal culture is relied upon exclusively.

The diagnosis of small bowel bacterial overgrowth in elderly patients

OBJECTIVE

To examine the frequency of small bowel bacterial overgrowth in elderly subjects.

DESIGN

Prospective observational study.

SETTING

Department of Health Care of the Elderly in a university teaching hospital.

PATIENTS

Thirteen elderly control subjects having undergone normal gastroscopy with normal hematology and anthropometry; 39 consecutive referrals of elderly patients with clinical or biochemical evidence of malnutrition.

MAIN OUTCOME MEASURES

Duodenal bacterial counts and the 14C-glycocholate, hydrogen, and 14C-xylose breath tests.

RESULTS

Five of the 13 control subjects and 21 of 39 patients with malnutrition had high duodenal bacterial counts (greater than 10(5) organisms/mL), whereas only seven patients were judged to have clinically significant bacterial overgrowth. The 14C-xylose breath test showed a high specificity (89%) but a low sensitivity (30%) when compared with the results of duodenal culture. The hydrogen breath tests and the 14C-glycocholate test showed a similar picture of moderately high specificity and low sensitivity (77% and 20%, respectively, for the hydrogen breath test, 76% and 33%, respectively, for the 14C-glycocholate test).

CONCLUSIONS

A positive xylose breath test was the best predictor of high duodenal bacterial counts. However, since many elderly subjects have high duodenal bacterial counts, the overdiagnosis of bacterial overgrowth syndrome will result if duodenal culture is relied upon exclusively.

Gallstone recurrence after medical dissolution. An overestimated threat?

We assessed gallstone recurrence rate in 42 patients diagnosed as having complete gallstone dissolution on bile acid therapy. By contrast with most previous studies, this diagnosis was based on ultrasound as well as on radiology, and only patients having their first gallstone recurrence were included in the study. Patients were followed for periods varying from 6 months to 7 years (median 30 months). Eleven patients had recurrences, giving an overall recurrence rate of 26%. A life analysis table was constructed by an actuarial method to compensate for the different lengths of follow-up in individual patients. Corrected recurrence rates by life table analysis were 15%, 21%, 25%, 36%, 45%, 45% and 45% at 1, 2, 3, 4, 5, 6 and 7 years respectively; for the same time intervals, cumulative recurrence rate overestimated the risk of gallstone recurrence (14%, 22%, 31%, 50%, 61%, 79% and 92%). We conclude that previous figures for recurrence rate have been an overestimate; but recurrence rate remains substantial over the first 5 years, and then levels off.

Gastric emptying and small-bowel transit rate in the elderly

Gastric emptying of liquid and small bowel transit in a group of elderly patients (mean age 79 years) was studied, with the elderly subjects compared to two groups of young controls. In 14 elderly patients, the initial rate of gastric emptying was significantly higher than in 14 young controls: the five-minute volume was 159 +/- 30 ml (mean +/- SEM) in the elderly patients and 294 +/- 22 ml in the young controls (p less than 0.005). The two groups' gastric emptying rates after five minutes were not significantly different. Small-bowel transit in 15 elderly patients was not significantly different from that found in 15 younger volunteers. These results suggest that gastric homeostatic mechanisms are impaired in the elderly.

Diurnal variation in cholesterol saturation of gall-bladder bile

In order to determine whether a diurnal variation in cholesterol saturation index is present in gall-bladder bile, samples of bile were taken by nasoduodenal intubation and cholecystokinin infusion at 9 am after the conventional 12 hour fast, and also at 5 pm five hours after a meal containing no cholesterol or phospholipid. In healthy controls saturation index (mean +/- SEM) fell from 1.02 +/- 0.08 at 9 am to 0.86 +/- 0.08 at 5 pm (n = 8, p less than 0.05). In untreated cholesterol gall-stone patients saturation index fell from 1.30 +/- 0.07 to 1.04 +/- 0.07 (n = 8, p less than 0.05); on chenodeoxycholic acid 15 mg/kg/day it fell from 0.91 +/- 0.06 to 0.78 +/- 0.07 (n = 16, p less than 0.01). The degree of diurnal variation was similar in those taking chenodeoxycholic acid at bedtime and in those taking it at mealtimes. The 9 am sample was supersaturated in three non-responders (showing no evidence of gall stone dissolution on oral cholecystogram after at least six months treatment) and in four responders. The 5 pm sample was a better predictor of treatment failure, being supersaturated in all four non-responders but in only one out of the 12 responders (p less than 0.01).

Assessment of gall-bladder storage function in man

1. We have validated a scintiscanning method for measuring fasting-state gall-bladder (GB) filling in man. 99mTc-labelled diethyl phenylcarbamoylmethyliminodiacetate (Tc-HIDA) was given intravenously, and 90 min later GB and gut activity were measured by using two isosensitive rectilinear scanning heads (anterior and posterior). Studies with a phantom GB in vitro, and studies in man in vivo, showed that the maximum error due to differences in isotope depth was 8%, compared with 300% when only one head was used.

2. By combining this technique with measurement of biliary lipid concentrations of fasting-state GB bile obtained by nasoduodenal intubation and intravenous cholecystokinin infusion, we were able to measure for the first time the total mass of all three biliary lipids in the GB. GB bile samples obtained in this way were divided into three consecutive portions of equal size in order to assess GB mixing. Bile acid pool size was also measured by isotope dilution.

3. We studied 12 healthy non-obese men. Fasting-state GB filling over 90 min (mean ± sem) was 54 ±8%. Biliary lipid mass in GB was 4.9 ±0.5 mmol for bile acids (67 ± 5% of the total bile acid pool), 1.6 ±0.2 mmol for phospholipid and 0.5 ± 0.1 mmol for cholesterol. The three consecutive portions of fasting GB bile gave values of 1.05 ± 0.07, 1.05 ± 0.06 and 1.03 ±0.10 for cholesterol saturation index (SI) and 6.6 ±1.1, 7.4 ± 1.6 and 6.5 ± 1.0 for Tc-HIDA c.p.m. × 1000 per mmol of bile acids.

4. The SI of fasting-state GB bile was significantly correlated with fasting-state GB filling (r = 0.63; P &lt; 0.05). It was also correlated with cholesterol mass in GB (r = 0.64; P &lt; 0.05), but not with bile acid and phospholipid mass.

5. We conclude that: (a) valid measurements of GB filling can be made in man by a simple scintiscanning technique employing 99mTc-HIDA as a biliary marker; (b) biliary lipid mass can also be measured if GB bile is obtained; (c) SI in health is in part determined by the degree of fasting-state GB filling, and in part by cholesterol mass in GB; (d) fasting-state GB content is well mixed in health.

Effect of sodium acetyl salicylate on cholesterol saturation of fasting gall bladder bile, with and without chenic acid

The aims of this study were (a) to determine whether sodium acetyl salicylate by itself or in combination with chenodeoxycholic (chenic) acid would decrease the cholesterol saturation index (SI) of fasting gall bladder bile in man; and (b) to confirm prospectively that 8 mg kg-1 day-1 chenic acid given at bedtime together with a low cholesterol diet reduces gall bladder bile SI to a level where consistent gallstone dissolution would be expected. Seven patients were studied on each of the following regimens given in random order for 1 month each: bedtime chenic acid alone 8 mg kg-1 day-1; sodium acetyl salicylate alone 600 mg four times daily; bedtime chenic acid together with sodium acetyl salicylate; no treatment. Gall bladder samples were taken by nasoduodenal intubation at the end of each regimen and SI determined. SI (mean +/- s.e. mean) on low cholesterol diet with no drug treatment was 1.14 +/- 0.06. On bedtime chenic acid 8 mg kg-1 day-1 plus low cholesterol diet it fell to 0.83 +/- 0.03 (P less than 0.05). Sodium acetyl salicylate did not alter gall bladder bile SI. 95% confidence limits for the effect of sodium acetyl salicylate on SI were +0.03 and -0.05. We conclude that (a) sodium acetyl salicylate does not lower SI of gall bladder bile in man; (b) an adequate fall in SI for gallstone dissolution can be achieved with a reduced dose (8 mg kg-1 day-1) of chenic acid given at bedtime with a low cholesterol diet.

Factors affecting gall-stone dissolution rate during chenic acid therapy

Gall-stone dissolution rate was measured in 52 patients with radiolucent gall stones in a functioning gall bladder receiving chenic acid 15 mg/kg/day. Percentage reduction in gall-stone area at six months was inversely related to initial gall-stone diameter (n = 52; r = -0.53; p less than 0.001), and to cholesterol saturation index of fasting gall-bladder bile during treatment (n = 28; r = -0.61; p less than 0.001). The duration of treatment required for complete dissolution was directly related to initial gall-stone diameter (r = 0.47; p less than 0.02). We conclude that the rate of gall-stone dissolution depends not only on gall-stone size but also on the degree of unsaturation of gall-bladder bile achieved during chenic acid treatment.

Gallstone dissolution rate during chenic acid therapy. Effect of bedtime administration plus low cholesterol diet

We have studied the dissolution rate of radiolucent gallstones from validated measurements of gallstone area derived from standardized oral cholecystography in 48 patients receiving chenodeoxycholic (chenic) acid, 15 mg/kg/day. Patients were matched in pairs according to gallstone size and randomly allocated to one of two treatment groups. Group A was given chenic acid in three divided doses at mealtimes with an unrestricted diet (the conventional regimen), and group B was given bedtime chenic acid with a low cholesterol diet. In the 24 matched pairs, the mean reduction in gallstone area at 6 months on regimen A was 36% and on B 60%. Median reductions in area were 20% and 57%, respectively (P less than 0.02). Of these matched pairs, 20 were followed at 6 month intervals for 42 months. There was a significant correlation between percentage reduction in gallstone area at 6 months and length of treatment necessary for complete dissolution (P less than 0.001). Fifty percent achieved complete dissolution by 30 months on regimen A, and by 12 months on regimen B (NS). We conclude that gallstone dissolution rate during chenic acid treatment is approximately doubled by bedtime administration plus a low cholesterol diet, thus reducing the duration and cost of treatment.

Minimum effective dose of chenic acid for gallstone patients: reduction with bedtime administration and low cholesterol diet

The aim of this study was to determine whether bedtime administration and a low cholesterol diet reduce the minimum effective dose of chenodeoxycholic (chenic) acid, defined as the dose giving a mean cholesterol saturation index of 0.8. Dose response studies were carried out in 10 patients with radiolucent gallstones in a functioning gallbladder during three different treatment regimens. On each regimen, all patients received three different doses of chenodeoxycholic acid in random order for one month each. Bedtime chenic acid plus a low cholesterol diet gave the greatest reduction in saturation index. A significant dose/response relationship was found on each regimen. On the conventional regimen of mealtime chenic acid, the minimum effective dose was 14 mg/kg/day; on bedtime chenic acid it was 12.4 mg/kg/day; and on bedtime chenic acid plus low cholesterol diet it was further reduced to 8.4 mg/kg/day (p less than 0.01). There was a dose-related increase in bowel frequency, which was absent at 10.6 mg/kg/day and below. We conclude that administration of chenic acid at bedtime with a low cholesterol diet enables the minimum effective dose for gallstone dissolution to be approximately halved, thus preventing diarrhoea and reducing the cost of treatment.

Postprandial gall-bladder emptying in patients with gall stones

Gall-bladder emptying in response to a standard meal was assessed in 34 patients with radiolucent gall stones and 34 matched controls. Percentage gall-bladder emptying, derived from volume measurements made on standardised oral cholecystography, was significantly higher at 15 minutes in the patients than the controls (mean +/- SE of mean 38.0 +/- 3.7% v 28.0 +/- 3.8%). This difference was maintained at 30 and 60 minutes. It is concluded that postprandial gall-bladder emptying is increased in patients with cholesterol gall stones, and that this may be the cause of the small bile-acid pool found in these patients.

Gall-bladder sensitivity to cholecystokinin in patients with gall stones

Gall-bladder sensitivity to cholecystokinin (CCK) was determined by dynamic cholescintigraphy in 18 patients with radiolucent gall stones and 18 matched controls during an infusion of CCK in which the rate of infusion was increased. In 10 of the matched pairs the patient was more sensitive than the control, in one the control was more sensitive, and in seven no difference was detected (p = 0.012). It is concluded that patients with cholesterol gall stones have increased gall-bladder sensitivity to CCK, and that this may be important in the pathogenesis of this disease.