Impact of Long-Term Enteral Feeding Tubes on Aspiration Pneumonia in a Tertiary Care Centre in Saudi Arabia
Muneerah Albugami1, Yasmin Al Twaijri2, Habib Bassil1, Ulrike Laudon1, Abeer Ibrahim3, *, Mohamed El Karouri1, Abdulaziz Al Rashed1, Abdelazeim Elamin1, Ahmed Sabry1, Rania Abdelreheem3, Abdulwahab Motieb1, Ali Al Araj1, Reem Hawary4, Sawsan Al Balawi4
1Internal Medicine Department, King Faisal Specialist Hospital and Research Center, Riyadh, KSA
2Research Center, King Faisal Specialist Hospital and Research Center, Riyadh, KSA
3Internal Medicine Department, Faculty of Medicine, Alexandria University, Alexandria, Egypt
4Nutrition Services, King Faisal Specialist Hospital and Research Center, Riyadh, KSA
To cite this article:
Muneerah Albugami, Yasmin Al Twaijri, Habib Bassil, Ulrike Laudon, Abeer Ibrahim, Mohamed El Karouri, Abdulaziz Al Rashed, Abdelazeim Elamin, Ahmed Sabry, Rania Abdelreheem, Abdulwahab Motieb, Ali Al Araj, Reem Hawary, Sawsan Al Balawi. Impact of Long-Term Enteral Feeding Tubes on Aspiration Pneumonia in a Tertiary Care Centre in Saudi Arabia. American Journal of Internal Medicine. Vol. 3, No. 3, 2015, pp. 95-102. doi: 10.11648/j.ajim.20150303.13
Abstract: It is a retrospective Chart Study. The objectives of the study are (1) to determine the incidence of Aspiration Pneumonia (AP) before and after long term feeding tubes insertion in four types of feeding tubes: percutaneous endoscopic gastrostomy (PEG), percutaneous fluoroscopy gastrostomy (PFG), jejunostomy feeding tube (JFT) and nasogastric tube (NGT) ,(2) to find out associations between the incidence of AP in patient who have feeding tubes and age , gender, rate of feeding (continuous or boluses) ,type of formula of used feeding ,use of thickener during oral feeding , persons deliver feedings and family training how to feed patients. (3) Factors that influenced patients’ outcomes. The findings of the study are: (1) No difference in incidence of AP before and after tube insertion. Feeding tubes have limited medical benefits for AP prevention. (2)Rate of feeding either continuous or bolus increase the frequency of AP. (3)No associations between the incidence of AP and age, gender, type of formula, use of thickener during oral feeding, person deliver feedings and family training about method of feeding. (4) Old age is a poor prognostic factor and HHC follow up is a good prognostic factor for outcome. (5) AP increases a patient's hospital readmission and length of stay in the hospital. There is an urgent need to have alternative strategies to reduce the cost.
Keywords: Aspiration Pneumonia, Enteral Feeding Tube, Mortality, Percutaneous Fluoroscopic Gastrostomy, Percutaneous Endoscopic Gastrostomy, Gastrojejunostomy
Aspiration Pneumonia (AP) is defined as the inhalation of either oropharyngeal or gastric contents into the lower airways. This is affected by quantity and nature of the aspirated material, the frequency of aspiration, and the host factors that predispose the patient to aspiration1. Aspiration of bacteria from oral and pharyngeal areas causes bacterial pneumonia2. A 10-year review found a 93.5% increase in the number of hospitalized elderly patients diagnosed with AP3. The mortality rate varied from one study to another in the range 7.5% to 62%. Deaths from AP are increasing and are currently ranked 15th on the CDC list of common causes of mortality4. AP diagnosis was based on a clinical presentation consistent with pneumonia associated with a history of witnessed aspiration or risk factors for aspiration 5. Almost all patients who develop AP have one or more of the predisposing risk factors for aspiration. Feeding tubes do not completely prevent pneumonia, it is associated with a greater incidence of pneumonia and a higher mortality6,7. Considering how common the problem of AP is in older adults, the use of feeding tubes has continued to increase in patients at King Faisal Specialist hospital & Research Centre (KFSH&RC). There is a limited data about the use of long term feeding tubes in Saudi patients8,9,10,11.
The main objective of this study is: 1)To determine the incidence of AP before and after long term feeding tubes insertion. 2)To find out if there are associations between incidence of AP in patients have feeding tubes and age , gender, type of tube used , rate of feeding (continuous or boluses), type of formula of feeding ,used thickener during oral feeding , persons deliver feedings and family training how to feed patients. 3) Factors that influenced patients’ outcomes .To our knowledge, our study is the first study conducted in Saudi Arabia about impact of feeding tubes on AP.
This is a retrospective study of patients who were admitted with a diagnosis of AP to KFSH&RC from January 2002 – December 2007. Inclusion criteria: 1- adult patient (age >14 years old) , 2- patients need long term enteral feedings tube > 4 weeks, 3- feedings tube inserted at KFSH&RC, 4- aspiration confirmed either by swallowing assessment test or modified barium test or both. Exclusion criteria include patients need feedings tubes for short term 4 weeks or less because of acute illness e.g. postoperative, ICU patients and patient’s terminal illness required palliative care. The study was approved by Office of Research Affair (ORA) at KFSH&RC. The diagnosis of AP was based on history of witnessed aspiration or recurrent chocking, one or more of the following symptoms: cough with or without sputum, fever or hypothermia, chest examination and chest x-ray findings confirmed a new infiltration at the day of admission .Antibiotic treatment was started for all patients at the emergency department. The patient chart was analyzed using the following parameters: demographic data of the patients (age, sex), indications for feeding tubes, frequency of AP before and after feeding tubes insertion it is defined to be less than 5times in last two years before tube insertion and more than 5 times in two years after tube insertion, dysphagia assessment by swallowing assessment test and modified barium test, types of feeding tubes (NGT, PEG, PFG or JFT). Patients may have different types of feeding tubes but we record only the first long term feeding tube inserted. Reasons to keep patients on long-term NGT feeding, rate of feeding (continuous or boluses), type of formula, use of thickener if patient is still taking oral diet, persons deliver the feeds (member of family, hospital nurse, house maid, home nurse), family training how to feed patients (training by medical staff at hospital before discharge patients or home health care service (HHC) of the hospital). Outcome: mortality from AP and factors that influenced outcomes
3. Data Analysis
All the statistical analysis of data was done by using the software package SAS version 9.3 (Statistical Analysis System, SAS Institute Inc., Cary, NC, USA). Descriptive statistics for the continuous variables are reported as mean ± standard deviation and categorical variables are summarized as frequencies and percentages. Continuous variables are compared by Student’s paired t-test while categorical variables are compared by Chi-square test. Univariate and multivariate logistic regression were used to study the effect of the different risk factors on the frequency of aspiration pneumonia after using the feeding tube and the patients’ outcome. The level of statistical significance is set at p < 0.05.
Numbers of patients were 389. Patients excluded from the study were 244 because of 227 patients had feeding tube inserted for short term which was less than 4 weeks, 7 patients their charts were missing and 10 patients their charts were at KFSH&RC- Jeddah and it was very difficult to be requested. Patients met the criteria of study are 145 , the main patients characteristics are presented in table 1, (83 men and 62 women) males are predominant (57.24%), with a mean age of 65.3 , bedridden patients are 85.03%, 21.38% have tracheostomy and 37.24% have follow up with HHC of the hospital. 14.97 % of patients are diabetic.The commonest indications for long term feeding tube are cerebrovascular accident (CVA) 49%, dementia 38.1%, inadequate oral intake 17.69% , it is not clear what is the underlying cause . Parkinson’s disease 6.1% and it is not documented if it is associated with dementia or not. Cancer patients are all in remission and no evidence of active disease as presented in table 1.Patients have two or more indications for feeding tubes are 29%. Swallowing assessment test is positive in 62.50% and modified barium swallow test (MBS) is positive in 55.10%. Swallowing assessment test and MBS are positive in 48.30% with p value<0.0001. The combination of two tests increased the diagnostic sensitivity to identify patients with silent aspirations. The commonest feeding tube used is PFG in 56.55% of patients as presented in table 1. The frequency of AP <5 times / year is 26.73% before the tube insertion and 90.91% after the tube insertion. The frequency of AP >5 times/ year is 73.3 % before the tube insertion and 9.09 % after the tube insertion. No difference in incidence of AP before and after tube insertion (p= 0.087) as presented in table 2. AP frequency is more in PFG but it could be because it is the commonest tube used, however, there is no difference between four types of feeding tubes in incidence of AP before and after the tube insertion (p =0.2331) as presented in table 3.The frequency of AP is more in male before and after the tube insertion , however it is statistically insignificant (p= 0.9795 and p= 0.5207) as presented in table 3.The frequency of AP is more in age 66-79 years before and after the tube insertion (p= 0.483) . The commonest type of formula used is Jevity 56.46% and there are no associations between incidence of AP and types of formula as presented in table 4. Thickener used in 9.52 % of patients and it has no effect on the incidence of AP (p= 0.1231) as presented in table 4. The rate of feeding either continuous or bolus increased the incidence of AP (p= 0.0318 and 0.0315) respectively. When we used univariate logistic regression to compare continuous and bolus of feeding, there is no difference and we conclude that they have the same effect. The persons deliver feedings are member of family38.36%, home nurse 10.96%, housemaid 9.59 % and hospital nurse6.85%. There are no associations between frequency of AP and person delivers feedings. Family training how to feed patients documented in 83 patients (56.85%). There is no association between frequency of AP and family training
(p= 0.554). Neither training by medical staff at hospital before discharge patient (p =1.056) nor HHC (p =0.081) nor both (p =0.456). The incidence of AP in patients who had follow up with HHC was small as compared to patients without HHC follow up but statistically it is insignificant (p 0.0561) as in figure 1. The survival rate is 43.4% and the mortality rate is 56.59% over the study period. The commonest cause of death is AP with septic shock and respiratory failure in 26 patients (37.68 %), followed by septic shock in 20 patients (28.99%), it is not clear if the cause is AP or other causes of sepsis. Malignancy was in 7 patients (10.14%), gastrointestinal bleeding with shock in 2 patients (2.90%). PFG has the highest mortality rate and it may be because it is the commonest feeding tube used among our patients as presented in table 5-6. Old age is a poor prognostic factor (p= 0.0018, odds ratio 1.028) and HHC follow up is a good prognostic factor for outcome, survival was better for patients have HHC follow up (p <0.0001, odds ratio 7.329). By using univariate and multiviate models we found old age and HHC follow up are the most significant prognostic factors, age (p= 0.0067, odds ratio 1.030), and HHC (p= <0.0001, odds ratio 8.379), however there are no association between outcome and gender (p =0.1776, odds ratio 0.614), dementia (p= 0.8254, odds ratio 1.085) and having two or more indications for feeding tubes insertion (p 0.5582, odds ratio 0.797). PFG has the highest number of death 40 (60.61%) followed by NGT14 (21.21%). Mean of length of hospital admission for AP after tube insertion per year is 22.1 days. The average cost of patient admission to medical floor per day at KFSH&RC is around 2882SR (768.5 $) in 22 days it will be 63404 SR (16907.7$). AP increases a patient's hospital readmissions, hospital stays and cost.
< 5times in last two years before tube insertion
> 5 time two years after tube insertion
|Mean age 65.3 ± 23.7|
|Tracheostomy when feeding tube inserted||31||21.38%|
|Diabetic patients||22||14.97 %|
|Indication for enteral feeding tube inserted|
|cerebrovascular accident (CVA)||74||51%|
|Inadequate oral intake||26||17.69%|
|Myopathy ,sever dysphagia||2||4.88%|
|Arnold chiari malformation||1||2.44%|
|Multiple systemic atrophy||1||2.44%|
|Childhood spinal atrophy||1||2.44%|
|Becker’s muscular dystrophy/multiple sclerosis||1||2.44%|
|Pituitary maroadenoma with hydrochelus||1||2.44%|
|Amyotrophic latral sclerosis||1||2.44%|
|Post brain tumor resection||1||2.44%|
|Low grade oligodendroglioma||1||2.44%|
|Cancer of tonsil||1||2.44%|
|Hypopharyngeal squamous cell carcinoma||1||2.44%|
|Patients have 2> indications for enteral feeding tube inserted||43||29.66%|
|Swallowing assessment test - Positive test||90||62.50%|
|Modified barium swallow test - Positive test||81||55.10%|
|types of feeding tubes|
|Jejunostomy feeding tubes (JFT)||13||8.97|
|percutaneous endoscopic gastrostomy (PEG)||19||13.10|
|percutaneous fluoroscopy gastrostomy ( PFG)||82||56.55|
|Home health care follow up||54||37.24%|
|Mean of length of Hospital admission for AP after tube insertion per year = 22.1|
|Frequency of hospital admission for AP||Before feeding tube insertion||After feeding tube insertion||P value|
< 5times in last two years before tube insertion , > 5 time two years after tube insertion
|Incidence of AP||NGT||JFT||PEG||PFG||P value|
|Before feeding tube|
|after feeding tube insertion|
Jejunostomy feeding tubes (JFT), percutaneous endoscopic gastrostomy (PEG), percutaneous fluoroscopy gastrostomy (PFG)
|Frequency||%||P value||Odds Ratio||95% confidence limits|
|Rate of feeding|
|Continuous||36||24.49%||0.0318||2.435||1.081 - 5.458|
|Bolus||86||58.50%||0.0315||0.435||0.204 – 0.929|
|Compare Continuous and Bolus||0.2911||1.711||0.631 – 1.461|
|0.2455||0.576||0.227 – 1.461|
|Type of formula|
|Jevity||83||56.46%||0.0632||0.489||0.230 – 1.040|
|Plumocare||5||3.40 %||0.0946||4.765||0.764 – 29.708|
|Insure||14||9.52 %||0.3339||0.467||0.099 – 2.190|
|Glucerna||22||14.97 %||0.1944||1.892||0.722 – 4.954|
|Nepro||3||2.04 %||0.1402||6.227||0.548 – 70.755|
|Suplena||4||2.72 %||0.2687||3.086||0.419 – 22.725|
|oral diet after tube insertion|
|Puried||23||15.65 %||0.9122||1.059||0.383 – 2.923|
|Liquid||3||2.05%||0.9803||< 0.001||< 0.001- >999.999|
|Regular||3||2.05%||0.9804||< 0.001||< 0.001- >999.999|
|Thickener used||14||9.52 %||0.1231||0.278||0.055 – 1.415|
|Person deliver feedings|
|Member of family||56||38.36%||0.2143||0.602||0.270 – 1.341|
|Home nurse||16||10.96%||0.9700||<0.001||< 0.001- >999.999|
|Housemaid||14||9.59 %||0.1336||0.205||0.026 – 1.625|
|Hospital nurse||10||6.85%||0.6887||0.722||0.146- 3.561|
|Family training||83||56.85%||0.1235||0.554||0.261 – 1.175|
|Family education by medical staff at hospital before discharge||51||34.93%||0.8918||1.056||0.481 – 2.318|
|Family education by HHC||5||3.42 %||0.800||0.081||0.081 – 6.937|
|Both||50||34.25 %||0.0785||0.456||0.190 – 1.094|
|follow up with Home health care||36||25%||14||9.72%|
|no follow up with Home health care||20||13.89%||57||39.58 %|
|Cause of death||JFT||NGT||PEG||PFG||total|
|Acute renal failure||0||0||0||1||1 (1.52%)|
|Heart failure||0||0||0||1||1 (1.52%)|
|Septic shock||3||3||4||10||20 (30.30%)|
|Septic shock + Acute renal failure||0||0||0||1||1 (1.52%)|
|Septic shock+ gastrointestinal bleeding||0||0||0||2||2 (3.03%)|
|Septic shock + Acute respiratory failure/acute respiratory distress syndrome + Aspiration pneumonia||1||6||2||16||25 (37.88%)|
In our study the incidence of AP is 62 % before tube insertion and 76 % after tube insertion. It is similar to other study finding of 64.3%12. The commonest indication for feeding tube in our study is CVA. Dysphagia after a stroke is the most common cause of PEG tubes insertion in more than 121,000 Medicare recipients in the United States 1. A systematic review reported that stroke patients with dysphagia demonstrate ≥3-fold increase in pneumonia risk with an 11-fold increase in pneumonia risk among patients with confirmed aspiration. Pneumonia is accounting for nearly 35% of post-stroke deaths13.The second commonest indication in our study is dementia. Despite lack of evidence that feeding tubes benefit patients with dementia, patients with dementia who have difficulty swallowing or reduced food intake often receive feeding tubes14. In US nursing homes, one third of residents with advanced dementia are tube fed15. We did not find any association association between AP incidence and patients have two or more indications for feeding tube. The commonest feeding tube used in our study is PFG. Previous studies found that PFG has proved to be efficient and safe: the rate of successful tube placement is 98% to 100%; PFG has a slightly higher success rate compared with PEG16.
In our study 20.69 % of patients had long term NGT because family refused to insert other types of feeding tubes. They don’t want to expose patients to any invasive procedures and thought that may be the patients will get better. JFT is uncommon to be used among our patients and it is not clear if there was a trial to insert other types of tubes before decision was made to insert JFT.
In our study there is no difference in incidence of hospital admission for AP before and after tube insertion. The frequency is more in PFG but it could be because it is the commonest tube used. When we compared the incidence of readmissions among different types of tubes, it is statistically insignificant. This enforces the facts that while feeding tubes are initiated to prevent AP, it does continue to occur. It was reported that the aspiration of oropharyngeal contents will continue and the risk of pneumonia remains high in patients on feeding tube17. In other studies, age and demented nursing home patients on long-term enteral feeding experienced significantly more episodes of AP compared with those nursing home patients who were not tube fed18. The literature review of the effect of feeding tubes in AP showed variation in rate of AP. Incidence of AP is 22.9% in gastrostomy tube fed nursing home patients in a retrospective review19 and 15.9% in jejunostomy-fed patients, so jejunostomy feedings do not offer effective protection against AP20. PEG was associated with a lower incidence of AP as compared to NGT21, 22. Direct percutaneous endoscopic jejunostomy (D-PEJ) was associated with lower incidence of AP as compared to percutaneous endoscopic gastrostomy-jejunostomy (PEGJ)23. AP occurs less frequently with PFG than with PEG24. However most of studies showed no difference between the different feeding tubes which is similar to our findings25,26,27,28,29,30,31.
The commonest type of formula used was Jevity. It is a calorically dense formula that has unique fiber blends which provides balanced and complete nutrition. It helps patients to maintain their weight. Glucerna is used in 22 patients (14.97 %) it is a reduced-carbohydrate, modified-fat, fiber-containing formula designed for people with diabetes. This means 14.97 % of patients in the study were diabetic. There are no associations between incidence of AP and type of formula and any type of feeding tubes. Thickener used has no effect on the incidence of AP. There is no relationship between incidence of AP and type oral diet on long-term enteral feeding patients. The reason of combination of oral diet and feeding tubes because of inadequate oral intake which documented in 17.69% of patients based on calories counting. The use of thickened liquids is one of the most frequently used compensatory interventions in hospitals and long-term care facilities. Only little evidence suggests that thickened liquids result in significant positive health outcomes with regards to nutritional status or pneumonia. Despite the overall lack of evidence supporting the use of thickened liquids, this strategy continues to be a cornerstone in dysphagia management in many facilities13.
We found a connection between incidence of AP and rate of feeding either continuous or bolus. Both have increased frequency of AP and have the same effect. This is similar to the finding of three randomized trials compared the two approaches and found that they have the same effect32, 33, 34 which is similar to other studies35, 36.
In our study, there are no relationships between the incidence of AP and person delivers feedings .This did not change whether training was done by medical staff at hospital before discharge patient or HHC of the hospital or both. Other studies have shown that by the time of discharge, caregivers should be adequately trained on the various aspects of the tube feeding system, to ensure safe and effective feeding at home 37. Interestingly, we found that the incidence of AP in patients who were followed up with HHC after discharge was less than patients without follow up.
To our knowledge, few studies have described the survival rate with such a long-term follow-up. Survival rate of patients have follow up with HHC was better as compared to no follow up with HHC. Over 6 years in our study, 56 patients (43.4%) survived. Long-term survival of geriatric patients in Japan treated with PEG showed 75% survived more than 6 months; 66% survived more than 1 year38. Others, found, survival after PEG insertion at 1, 6, 12, and 24 months were 90.5%, 52%, 42%, and 35%, respectively39. Other study of 68 cases (88%) showed that the 1-year survival rate was 64.0%, and the 2-year survival rate was 55.5 %40. Patients who receive a percutaneous feeding tube have a 30-day mortality risk of 18%–24% and a 1-year mortality risk of 50%–63%41. The largest report focused on 80,000 Medicare patients who had undergone PEG or surgical gastrostomy, the overall in-hospital mortality rate was 15%. In other report mortality at one and three years was 63 and 81 %, respectively42.
In our study the commonest cause of death is AP with septic shock and respiratory failure. PFG has highest mortality rate may be because it is the commonest feeding tube used among our patients.
Old age is a poor prognostic factor associated with a higher mortality (p= 0.0018, odds ratio 1.028) and survival was better for patients have HHC follow up (p <0.0001, odds ratio 7.329). By using multiviate model we found age and HHC follow up are the most significant prognostic factors, age (p= 0.0067, odds ratio 1.030), and HHC (p <0.0001, odds ratio 8.379), however there are no association between outcome and gender (p= 0.1776, odds ratio 0.614), dementia (p= 0.8254, odds ratio 1.085) and having two or more indications for feeding tubes insertion (p= 0.5582, odds ratio 0.797).No randomized clinical trials (RCTs) have been done about enteral tube feeding , considerable evidence from studies of weaker design strongly suggest that tube feeding does not reduce the risks of death, AP, pressure ulcers, other infections, or poor functional outcome 14, 43.
Mean of length of hospital admission for AP after tube insertion per year is 22.1. AP increases a patient's hospital readmissions, the length of stay in the hospital is long and the cost is high. There are limited data on the economic costs of patient’s hospital readmissions due to AP. The cost of managing a patient with a feeding tube (PEG) is reported to average over $31,000 per patient per year. The main components of this cost include the initial PEG procedure, enteral formula, and hospital charges for major complications44. There is urgent need to have strategies to reduce the cost.
This study has some limitations. It is a retrospective chart review where some missing data are expected and poor documentation was common during data collections. Since this study was performed at tertiary care hospital, generalizability may be limited due to small sample size. However, the size and diversity of the patient sample should help to reduce the potential effects of that limitation. Despite these limitations, this study finding is: (1) Comparison of four types of long-term enteral feeding showed no difference in incidence of AP before and after tube insertion. (2) Feeding tubes have limited medical benefits for AP prevention. Rate of feeding either continuous or bolus increase the frequency of AP. (3) No associations between the incidence of AP and age, gender, type of formula , thickener used, person deliver feedings and family training about method of feeding. (4) Old age is a poor prognostic factor and HHC follow up is a good prognostic factor for outcome. (5) AP increases a patient's hospital readmissions and length of stay in the hospital. There is an urgent need to have alternative strategies to reduce the cost.