Advertisement

Reduced Pneumonia Risk in Advanced Dementia Patients on Careful Hand Feeding Compared With Nasogastric Tube Feeding

Open AccessPublished:April 27, 2022DOI:https://doi.org/10.1016/j.jamda.2022.03.011

      Abstract

      Objectives

      To compare survival and pneumonia risk among hospitalized patients with advanced dementia on nasogastric tube feeding (NGF) vs careful hand feeding (CHF) and to examine outcomes by feeding problem type.

      Design

      Retrospective cohort study.

      Setting and Participants

      Advanced dementia patients aged ≥60 years with indication for tube feeding admitted to 2 geriatric convalescent hospitals between January 1, 2015, and June 30, 2019.

      Methods

      Comparison on the effect of NGF and CHF on survival and pneumonia risk using Kaplan Meier survival analysis and Cox proportional hazards models.

      Results

      Of the 764 patients (mean age 89 years, 61% female, 74% residential care home residents), 464 (61%) were initiated on NGF and 300 (39%) on CHF. The primary feeding problem types were dysphagia (50%), behavioral feeding problem (33%), or both (17%). There was no difference in 1-year survival rate between NGF and CHF groups (36% vs 37%, P = .71) and survival did not differ by feeding problem type. Nasogastric tube feeding was not a significant predictor for survival (adjusted hazard ratio 1.15, 95% CI 0.94-1.39). Among 577 (76%) patients who survived to discharge, pneumonia rates were lower in the CHF group (48% vs 60%, P = .004). After adjusting for cofounders, NGF was a significant risk factor for pneumonia (adjusted hazard ratio 1.41, 95% CI 1.08-1.85). In subgroup analyses, NGF was associated with increased pneumonia risk for patients with both dysphagia and behavioral feeding problem (P = .01) but not in patients with behavioral feeding problem alone (P = .24) or dysphagia alone (P = .30).

      Conclusions and Implications

      For advanced dementia patients with feeding problems, there is no difference in survival between NGF and CHF. However, NGF is associated with a higher pneumonia risk, particularly for patients with both dysphagia and behavioral feeding problem. Further research on how the feeding problem type impacts pneumonia risk for patients on NGF is needed.

      Keywords

      Feeding problems are common in patients with advanced dementia and can indicate progression of the disease toward the end of life.
      • Mitchell S.L.
      • Teno J.M.
      • Kiely D.K.
      • et al.
      The clinical course of advanced dementia.
      Feeding tubes are commonly inserted because of perceived benefits of reducing risk of aspiration pneumonia and improving survival.
      • Li I.
      Feeding tubes in patients with severe dementia.
      • Mitchell S.L.
      • Lawson F.M.
      Decision-making for long-term tube-feeding in cognitively impaired elderly people.
      • Carey T.S.
      • Hanson L.
      • Garrett J.M.
      • et al.
      Expectations and outcomes of gastric feeding tubes.
      • Hanson L.C.
      • Garrett J.M.
      • Lewis C.
      • et al.
      Physicians' expectations of benefit from tube feeding.
      However, prior observational studies have not found conclusive evidence that feeding tubes prevent aspiration pneumonia or prolong survival in patients with advanced dementia.
      • Finucane T.E.
      • Christmas C.
      • Travis K.
      Tube feeding in patients with advanced dementia: A review of the evidence.
      • Sampson E.L.
      • Candy B.
      • Jones L.
      Enteral tube feeding for older people with advanced dementia.
      • Lee Y.F.
      • Hsu T.W.
      • Liang C.S.
      • et al.
      The efficacy and safety of tube feeding in advanced dementia patients: A systemic review and meta-analysis study.
      In the last decade, several professional societies have issued guidelines that recommended careful hand feeding (CHF) as an alternative to tube feeding given the lack of benefit and the associated treatment burdens.
      American Geriatrics Society Ethics CommitteeClinical Practice and Models of Care Committee
      American Geriatrics Society feeding tubes in advanced dementia position statement.
      • Volkert D.
      • Chourdakis M.
      • Faxen-Irving G.
      • et al.
      ESPEN guidelines on nutrition in dementia.
      • Allan R.
      Oral feeding difficulties and dilemmas–a working party report.
      Australian and New Zealand Society for Geriatric MedicineAustralian and New Zealand Society for Geriatric Medicine
      Position statement–dysphagia and aspiration in older people.
      Despite these guidelines, controversy regarding the evidence on tube feeding in advanced dementia remains.
      • Delegge M.H.
      Percutaneous endoscopic gastrostomy in the dementia patient: helpful or hindering?.
      ,
      • Lynch M.C.
      Is tube feeding futile in advanced dementia?.
      A recent Cochrane review found a high risk of bias due to confounding and selection bias among existing controlled, nonrandomized studies.
      • Davies N.
      • Barrado-Martín Y.
      • Vickerstaff V.
      • et al.
      Enteral tube feeding for people with severe dementia.
      Critics challenged whether patients on tube feeding and those on oral feeding are comparable groups as the oral feeding patients may not have an indication for tube feeding.
      • Lynch M.C.
      Is tube feeding futile in advanced dementia?.
      ,
      • Davies N.
      • Barrado-Martín Y.
      • Vickerstaff V.
      • et al.
      Enteral tube feeding for people with severe dementia.
      Furthermore, the majority of studies were on percutaneous endoscopic gastrostomy (PEG) feeding or did not differentiate between nasogastric tube feeding (NGF) and PEG feeding.
      • Lee Y.F.
      • Hsu T.W.
      • Liang C.S.
      • et al.
      The efficacy and safety of tube feeding in advanced dementia patients: A systemic review and meta-analysis study.
      ,
      • Davies N.
      • Barrado-Martín Y.
      • Vickerstaff V.
      • et al.
      Enteral tube feeding for people with severe dementia.
      Two studies that focused on NGF showed mixed results on mortality risk.
      • Alvarez-Fernández B.
      • García-Ordoñez M.A.
      • Martínez-Manzanares C.
      • Gómez-Huelgas R.
      Survival of a cohort of elderly patients with advanced dementia: Nasogastric tube feeding as a risk factor for mortality.
      ,
      • Chou H.H.
      • Tsou M.T.
      • Hwang L.C.
      Nasogastric tube feeding versus assisted hand feeding in-home healthcare older adults with severe dementia in Taiwan: A prognosis comparison.
      Additional studies on NGF are needed given this is the predominant type of enteral feeding used in people with dementia in many Asian countries including China, Taiwan, and Singapore.
      • Zhang Y.
      • Ma C.
      • Li C.
      • Chen Q.
      • Shen M.
      • Wang Y.
      Clinician's attitude to enteral nutrition with percutaneous endoscopic gastrostomy: A survey in China.
      • Wong A.
      • Sowa P.M.
      • Banks M.D.
      • Bauer J.D.
      Home enteral nutrition in Singapore's long-term care homes—incidence, prevalence, cost, and staffing.
      In Hong Kong, a study reported that the prevalence of NGF was 53% among residents with advanced cognitive impairment in 66 residential care homes.
      • Luk J.K.H.
      • Chan W.K.
      • Ng W.C.
      • et al.
      Mortality and health services utilization among older people with advanced cognitive impairment living in residential care homes.
      Following the release of the 2015 Hospital Authority Guidelines on Life-Sustaining Treatments in the Terminally Ill, which advocated a palliative care approach in managing feeding problems in advanced dementia,
      some geriatric convalescent hospitals began to shift from the default practice of nasogastric tube insertions to discussing careful hand feeding with family surrogates as an alternative.
      • Luk J.K.H.
      • Chan T.C.
      • Chan F.H.W.
      Careful hand feeding program in a geriatric step-down hospital in Hong Kong–is this feasible?.
      Given the majority of feeding tube insertions occur during hospitalization,
      ,
      • Kuo S.
      • Rhodes R.L.
      • Mitchell S.L.
      • et al.
      Natural history of feeding-tube use in nursing home residents with advanced dementia.
      hospitals that offer the choice of CHF provide opportune settings for clinically meaningful comparison of the effects of different feeding modes in advanced dementia patients with feeding problems.
      We conducted a 12-month retrospective longitudinal study of hospitalized patients with advanced dementia who were initiated on NGF vs CHF at 2 geriatric convalescent hospitals that support CHF practices in Hong Kong. The primary objective was to compare the effects of NGF vs CHF on survival and pneumonia risk on hospitalized advanced dementia patients with feeding problems. The secondary objective was to examine whether the impact of feeding mode on survival and pneumonia risk differed by feeding problem type.

      Methods

      Study Design

      This was a retrospective cohort study. Data were collected from the Hong Kong Hospital Authority paper admission charts and electronic medical record system, which links medical records for all public hospitals in Hong Kong.

      Setting and Participants

      Medical charts of patients who underwent a speech therapist evaluation during admission to 2 geriatric convalescent hospitals in Hong Kong between January 2015 to June 2019 were reviewed.
      Eligibility criteria included age ≥60 years, advanced dementia defined as stage 7 on the Reisberg Global Deterioration Scale (GDS),
      • Reisberg B.
      • Ferris S.H.
      • de Leon M.J.
      • Crook T.
      The global deterioration scale for assessment of primary degenerative dementia.
      and documented indication for tube feeding during the hospitalization due to at least 1 of the following feeding problems: poor oral intake or aspiration risk due to dysphagia. Patients were categorized into 2 cohorts, NGF or CHF group, based on the consensus feeding mode decision made between the treating physician and the family surrogate.
      Patients were excluded if they had (1) any type of feeding tube in place at the time of hospital admission, (2) acute stroke, (3) tracheostomy, (4) active cancer, (5) were comatose, or (6) missing data on the type of feeding problem.
      Three speech therapists performed bedside swallowing evaluations and assessed the type and severity of feeding problems. The feeding problems were categorized as (1) oropharyngeal dysphagia, (2) behavioral feeding problem, or (3) presence of both types. Behavioral feeding problem refers to difficulty getting food into the mouth or behaviors such as food refusal or spitting out food, leading to poor oral intake.
      • Volicer L.
      • Seltzer B.
      • Rheaume Y.
      • et al.
      Eating difficulties in patients with probable dementia of the Alzheimer type.
      • Stockdell R.
      • Amella E.J.
      The Edinburgh Feeding Evaluation in Dementia Scale: determining how much help people with dementia need at mealtime.
      • Chan C.P.H.
      • Kwan Y.K.
      Feeding-swallowing issues in older adults with dementia.
      The degree of swallowing disability was rated using the validated Royal Brisbane Hospital Outcome Measure for Swallowing.
      • Ward E.C.
      • Conroy A.L.
      Validity, reliability and responsivity of the Royal Brisbane Hospital outcome measure for swallowing.
      Regular internal audits on dysphagia management were conducted to align the practice of the speech therapists in the 2 hospitals.
      For patients initiated on CHF, a speech therapist provided an individualized written CHF plan with recommendations on food consistency, feeding techniques, and utensils. A dietician conducted a nutritional assessment and provided oral supplementation if needed. The physician assessed the patient's medical and psychiatric conditions and adjusted any medications to improve appetite, gastric emptying, and sensorium. The CHF plan was implemented and periodically reviewed by the multidisciplinary team consisting of the physician, nurse, speech therapist, and dietician. The patient's primary caregiver was permitted to provide feeding assistance to the patient if desired and was taught to observe for signs of choking and aspiration. The CHF plan was provided to the primary caregiver at discharge.

      Data Collection

      To examine whether the type and severity of feeding problem affected survival and pneumonia risk in patients on NGF or CHF, we collected data on the speech therapist's bedside evaluation of the feeding problem as the primary indication for tube feeding.
      Based on previous studies,
      • Meier D.E.
      • Ahronheim J.C.
      • Morris J.
      • Baskin-Lyons S.
      • Morrison R.S.
      High short-term mortality in hospitalized patients with advanced dementia: Lack of benefit of tube feeding.
      • Teno J.M.
      • Gozalo P.L.
      • Mitchell S.L.
      • et al.
      Does feeding tube insertion and its timing improve survival?.
      • Ticinesi A.
      • Nouvenne A.
      • Lauretani F.
      • et al.
      Survival in older adults with dementia and eating problems: To PEG or not to PEG?.
      the following potential confounding variables were collected: sociodemographic variables including age, sex, marital status, education, and type of residence (home vs residential care home); nutritional parameters including body mass index (BMI), serum albumin, and lymphocyte count; presence of pressure injury; measures of functional status including Functional Assessment Staging Tool and level of dependence for activities of daily living; aspiration pneumonia prior to feeding mode decision; history of stroke and other comorbidities for determining the Charlson Comorbidity Index (CCI); and evidence of advance care planning including advance care planning documentation and Do-Not-Attempt Cardiopulmonary Resuscitation (DNACPR) orders for nonhospitalized patients.

      Outcome Measures

      The primary outcomes were 1-year survival and pneumonia incidence. Survival was measured as the number of days between feeding mode decision and date of death. Pneumonia was based on the treating physician's clinical diagnosis of new occurrence of pneumonia after the feeding mode decision.

      Statistical Methods

      Independent samples t test and chi-square test were used for comparing continuous and categorical variables between the 2 cohorts. Survival and time to pneumonia was estimated by using the Kaplan-Meier method and differences between groups was compared using log-rank test. All analyses were conducted based on the original feeding mode assigned, regardless of whether the feeding mode was changed during the follow-up period. Predictors on survival and pneumonia risk were first assessed using univariate Cox proportional hazards regression models. Risk factors with a P < .1 were included in the multivariate model. For all other outcomes, P < .05 was considered to indicate statistical significance. Statistical analysis was conducted using SPSS, version 26. The local institutional review board approved this research project.

      Results

      Characteristics of the Subjects

      Among the 764 patients, 464 (60.7%) were initiated on NGF and 300 (39.1%) on CHF. Baseline characteristics of participants are presented in Table 1. The NGF group was younger than the CHF group (88.2 vs 89.8 years, P < .001). There were no differences in sex, education levels, and place of residence. Compared with the CHF group, the NGF group were more likely to have dysphagia or both dysphagia and behavioral feeding problem, higher severity of dysphagia, and a diagnosis of aspiration pneumonia. The NGF group also had more patients with complete activities of daily living dependence than the CHF group. Both groups were similar in nutritional status, Charlson Comorbidity Index, FAST level, and the presence of pressure injury. The CHF group had more advance care planning and DNACPR documentation prior to the hospitalization.
      Table 1Characteristics of Participants on NGT vs CHF
      CharacteristicNGF (n = 464)CHF (n = 300)Total (N = 764)P Value
      Age, y, mean (SD)88.2 (7.5)89.8 (7.1)88.8 (7.4)<.001
      Female sex280 (60.3)187 (62.3)467 (61.1).58
      Marital status.03
       Married148 (32.7)87 (29.5)235 (31.4)
       Single34 (7.5)9 (3.1)43 (5.7)
       Divorced or separated11 (2.4)6 (2.0)17 (2.3)
       Widowed260 (57.4)193 (65.4)453 (60.6)
      Education.78
       Illiterate255 (58.1)161 (55.7)416 (57.1)
       Primary126 (28.7)88 (29.5)214 (29.4)
       Secondary39 (8.9)30 (10.4)69 (9.5)
       Tertiary19 (4.3)10 (3.5)29 (4.0)
      Place of residence.78
       Residential care home346 (74.6)221 (73.7)567 (74.2)
       Home118 (25.4)79 (26.3)197 (25.8)
      Feeding problem<.001
       Behavioral97 (20.9)155 (51.7)252 (33.0)
       Dysphagia274 (59.1)111 (37.0)385 (50.4)
       Both93 (20.0)34 (11.3)127 (16.6)
      Severity of dysphagia<.001
       Mild41 (9.3)49 (17.3)90 (12.4)
       Mild-moderate18 (4.1)32 (11.3)50 (6.9)
       Moderate110 (24.9)82 (29.0)192 (26.5)
       Moderate-severe80 (18.1)32 (11.3)112 (15.4)
       Severe193 (43.7)88 (31.1)281 (38.8)
      Aspiration pneumonia270 (64.6)148 (35.4)418 (54.9).02
      Body mass index, mean (SD)18.6 (3.7)18.9 (4.2)18.7 (3.9).27
      Albumin, g/dL, mean (SD)28.2 (6.1)28.4 (6.2)28.3 (6.1).68
      Lymphocyte, 109 cells/L, mean (SD)1.1 (0.6)1.2 (1.2)1.2 (0.9).28
      Active pressure injury244 (52.7)151 (50.3)395 (51.8).52
      Comorbidities
       Myocardial infarction23 (5.0)12 (4.0)38 (5.0).79
       Chronic heart failure60 (13.0)33 (11.0)92 (12.0).44
       PVD14 (3.0)9 (3.0)23 (3.0).66
       CVA144 (31.0)93 (31.0)237 (31.0).96
       COPD19 (4.0)18 (6.0)38 (5.0).32
       Peptic ulcer9 (2.0)3 (1.0)8 (1.0).15
       Liver disease3 (0.7)2 (0.7)5 (0.7).97
       Diabetes mellitus130 (28.0)89 (29.7)219 (28.7).62
       Chronic kidney disease104 (22.5)81 (27.0)185 (24.3).15
      CCI, mean (SD)5.9 (1.2)5.9 (1.1)5.9 (1.1).97
      FAST level.13
       7A25 (5.4)16 (5.3)41 (5.4)
       7B50 (10.8)45 (15.0)95 (12.4)
       7C43 (9.3)37 (12.3)80 (10.5)
       7D-F346 (74.6)202 (67.3)548 (71.7)
      Functional status.02
       Partial ADL dependence27 (5.8)31 (10.3)58 (7.6)
       Complete ADL dependence437 (94.2)269 (89.7)706 (92.4)
      ACP
       ACP documentation9 (2.0)30 (10.0)38 (5.0)<.001
       Nonhospitalized DNACPR9 (2.0)39 (13.0)46 (6.0)<.001
      ACP, advance care planning; ADL, activities of daily living; CCI, Charlson comorbidity index; CHF, careful hand feeding; COPD, chronic obstructive pulmonary disease; CVA, cerebrovascular accident; DNACPR, do not attempt cardiopulmonary resuscitation; FAST, Functional Assessment Staging Tool; NGF, nasogastric tube feeding; PVD, peripheral vascular disease.
      Unless otherwise noted, data are presented as n (%).
      During the 1-year follow-up, 8 participants (1.7%) in the NGF group switched to careful hand feeding. In the CHF group, 29 (9.7%) were converted to nasogastric tube feeding during a subsequent hospitalization.

      One-Year Survival

      One-year mortality rate was 64.0% for the NGF group and 62.7% for the CHF group (P = .55). There was no difference in 1-year survival curves between the NGF and CHF groups, with median survival of 125 and 145 days, respectively (Figure 1A). Figure 1B–D shows the survival curves for patients on NGF and CHF by feeding problem type. No difference in survival was found between patients with behavioral feeding problem, dysphagia, or both feeding problem types.
      Figure thumbnail gr1
      Fig. 1Survival probability for advanced dementia patients on nasogastric tube feeding or careful hand feeding for (A) all patients, and by cohorts of primary feeding problem type: (B) behavioral feeding problem, (C) dysphagia, and (D) coexisting behavioral feeding problem and dysphagia.
      No association was found between feeding mode and survival in either univariate or multivariate Cox regression models (P = .18, adjusted hazard ratio 1.15, 95% CI 0.94-1.39) (Supplementary Table 1). Predictors of survival after multivariate analysis were age, marital status, diagnosis of aspiration pneumonia, albumin level, presence of pressure injury, and CCI (Table 2).
      Table 2Independent Predictors for Survival After Multivariate Analysis
      VariableP ValueHR (95% CI)
      Age<.0011.04 (1.03, 1.06)
      Marital status (referent: married).01
       Single<.0010.43 (0.26, 0.73)
       Divorced or separated.721.13 (0.59, 2.14)
       Widowed.140.83 (0.64, 1.07)
      Aspiration pneumonia<.0011.39 (1.14, 1.69)
      Albumin level<.0010.95 (0.94, 0.97)
      Active pressure injury.011.29 (1.07, 1.56)
      Charlson Comorbidity Index<.0011.19 (1.09, 1.29)
      HR, hazard ratio.

      Pneumonia Risk

      The pneumonia rate was lower in the CHF group (48.0%) than in the NGF group (60.3%) at 1 year (P = .004). The Kaplan-Meier plot for time to pneumonia for patients who survived to discharge are presented for the 2 feeding modes in Figure 2. NGF was a significant risk factor for pneumonia in the unadjusted and adjusted models (adjusted hazard ratio = 1.41, 95% CI 1.08-1.85) (Supplementary Table 2, Supplementary Table 3, Supplementary Table 4). Other significant predictors for pneumonia included gender, aspiration pneumonia prior to feeding mode decision, and BMI (Table 3). The severity of dysphagia was not an independent predictor of pneumonia risk in both the unadjusted and adjusted models. The feeding problem type was significant only in the unadjusted model but not in the adjusted model.
      Figure thumbnail gr2
      Fig. 2Pneumonia-free survival probability for advanced dementia patients on nasogastric tube feeding or careful hand feeding for (A) all patients, and by cohorts of primary feeding problem type: (B) behavioral feeding problem, (C) dysphagia, and (D) coexisting behavioral feeding problem and dysphagia.
      Table 3Independent Predictors for Pneumonia After Multivariate Analysis
      VariableP ValueHR (95% CI)
      Male.031.41 (1.04, 1.91)
      Nasogastric tube feeding (referent: CHF).011.41 (1.08, 1.85)
      Aspiration pneumonia.041.31 (1.01, 1.70)
      Body mass index.040.97 (0.93, 1.00)
      HR, hazard ratio; CHF, careful hand feeding.
      Subgroup analyses showed no significant difference in pneumonia risk between the NGF and CHF groups in patients with behavioral feeding problem alone (P = .33) or dysphagia alone (P = .89) (Figure 2B, C). NGF was associated with higher pneumonia risk in patients with coexisting dysphagia and behavioral feeding problem (P = .001) (Figure 2B).

      Discussion

      This study was the first to examine the effects of NGF vs CHF on survival and pneumonia risk in hospitalized advanced dementia patients and compare the effects by feeding problem type. The findings showed that NGF was not associated with improved survival, regardless of the feeding problem type and was associated with higher pneumonia risk, particularly in patients with coexisting behavioral feeding problem and dysphagia.
      Feeding mode decisions made by consensus between the treating physician and surrogate were influenced by both clinical factors and preferences of patients and surrogates. This was evidenced by higher rates of NGF among patients with dysphagia as the primary indication for tube feeding and higher severity of dysphagia. On the other hand, higher rates of CHF were found in patients with prior advance care planning and nonhospitalized DNACPR orders.
      Prior research suggested that feeding tubes did not confer survival benefit, but the certainty of evidence was low because of risk of selection bias and limited controlling for confounders.
      • Sampson E.L.
      • Candy B.
      • Jones L.
      Enteral tube feeding for older people with advanced dementia.
      ,
      • Davies N.
      • Barrado-Martín Y.
      • Vickerstaff V.
      • et al.
      Enteral tube feeding for people with severe dementia.
      The current study found a lack of survival benefit with NGF compared with CHF and attempted to reduce risk of bias by selecting hospitalized patients with indication for tube feeding due to feeding problems and excluded patients with existing feeding tubes. We also controlled for a wide range of potential confounders through comprehensive review of hospital paper charts and the electronic medical record. Another strength of this study was the identification of patients in hospitals that offered the option of CHF and NGF when real-time feeding tube decisions were made. This provided a clinically meaningful starting point for comparison of survival and pneumonia risk.
      Prior studies provided inconclusive evidence on the association between tube feeding and pneumonia risk.
      • Lee Y.F.
      • Hsu T.W.
      • Liang C.S.
      • et al.
      The efficacy and safety of tube feeding in advanced dementia patients: A systemic review and meta-analysis study.
      This study found that CHF was associated with lower pneumonia risk among patients who survived to discharge. In contrast, Chou et al
      • Chou H.H.
      • Tsou M.T.
      • Hwang L.C.
      Nasogastric tube feeding versus assisted hand feeding in-home healthcare older adults with severe dementia in Taiwan: A prognosis comparison.
      found no difference in pneumonia risk between NGF and assisted hand feeding patients receiving home-based health care in Taiwan. These varying results can be attributed to differences in the study populations whereby in the current study, the patients were hospitalized, primarily residential care home residents, and had greater than 5-fold higher 1-year mortality rate. There may also have been differences in the strategies used in CHF in this study and assisted hand feeding in the Chou et al study.
      The impact of the feeding problem type and severity of dysphagia on pneumonia risk had not been well studied in advanced dementia patients. Our analysis showed that the feeding problem type and the severity of dysphagia were not independent predictors of pneumonia. However, patients with higher severity of dysphagia were more likely to be initiated on NGF, and NGF was a significant risk factor for pneumonia.
      Dysphagia is a known risk factor for aspiration pneumonia, and the risk increases with the severity of swallowing dysfunction.
      • Payne M.
      • Morley J.E.
      Editorial: Dysphagia, dementia and frailty.
      We postulate that dysphagia and severity of dysphagia were not significant predictors of pneumonia in our study because the speech therapist's bedside evaluation may have reflected the patient's swallowing dysfunction at one point in time during the acute illness, and the degree of swallowing dysfunction could have changed over time in some cases. Nonetheless, the practice of initiating long-term NGF in hospitalized dementia patients considered to have high aspiration risk after a speech therapist's bedside assessment of dysphagia is not supported by the results of this study.
      In the current study, NGF was associated with increased pneumonia risk, particularly in patients with coexisting dysphagia and behavioral feeding problem and a nonsignificant trend that NGF confers higher pneumonia risk in patients with behavioral feeding problem only. This is a preliminary finding that needs to be further examined in a powered study to compare patients with different feeding problem types and objective measures of swallowing dysfunction.
      The postulated mechanisms for increased risk of aspiration pneumonia from NGF include impairment of lower esophageal sphincter with the presence of the tube across the gastric cardia, reduced frequency of esophageal body contractions, desensitization of the pharyngoglottal adduction reflex, and reflux of gastric contents into the pharynx, especially when patients are fed in a supine position.
      • Blumenstein I.
      • Shastri Y.M.
      • Stein J.
      Gastroenteric tube feeding: techniques, problems and solutions.
      ,
      • Gomes G.F.
      • Pisani J.C.
      • Macedo E.D.
      • Campos A.C.
      The nasogastric feeding tube as a risk factor for aspiration and aspiration pneumonia.
      On the other hand, we hypothesize that careful hand feeding may reduce the risk of aspiration pneumonia owing to modified food textures, erect positioning of the patient, and careful observations made by the carer during feeding, with cessation of feeding on noting the first signs of choking or respiratory distress. Additional studies are needed to examine the potential effects of feeding techniques and food modifications used in CHF on reducing aspiration pneumonia risk.
      This study has several limitations. First, a randomized controlled trial cannot be conducted because of ethical concerns. In this retrospective cohort study, there is a risk of selection bias. Although differences found in baseline demographic and clinical characteristics between the NGF and CHF groups were adjusted for in the Cox proportional hazards models, there may be other unmeasured differences between the 2 groups that could possibly have affected the choice of feeding mode. We did not investigate the effects of the type of dementia as data was frequently unavailable. Furthermore, this study concerned only hospitalized patients with advanced dementia and did not include patients on percutaneous gastrostomy tubes which were uncommonly used in our setting. Another limitation was the reliability and accuracy of the speech therapists' bedside swallowing evaluation cannot be assessed. Nonetheless, the speech therapists' evaluation reflects what was feasible in actual practice and influenced the feeding tube decisions in this study. Finally, we could not verify that the occurrence of pneumonia was due to aspiration pneumonia and acknowledge the possibility that other causes of pneumonia were included.

      Conclusions and Implications

      This study provides evidence that NGF does not confer survival benefit compared with CHF in patients with advanced dementia. On the other hand, CHF is associated with lower pneumonia risk compared with NGF. Future studies to examine how the feeding problem type affects pneumonia risk, particularly in patients on NGF, is warranted.
      This study has implications for clinicians, family and institutional caregivers, and hospitals that are involved in caring for dementia patients. Feeding problems are a natural part of the trajectory of advanced dementia and while feeding tube insertions are declining in some countries such as the United States,
      • Mitchell S.L.
      • Mor V.
      • Gozalo P.L.
      • et al.
      Tube feeding in US nursing home residents with advanced dementia, 2000–2014.
      the practice remains widespread in Hong Kong and many parts of world.
      ,
      • Ojo O.
      The challenges of home enteral tube feeding: A global perspective.
      ,
      • Best C.
      • Hitchings H.
      Enteral tube feeding–from hospital to home.
      The decision to insert feeding tubes is influenced by a myriad factors including local laws, cultural values, and institutional practices.
      • Volkert D.
      • Chourdakis M.
      • Faxen-Irving G.
      • et al.
      ESPEN guidelines on nutrition in dementia.
      ,
      • Mitchell S.L.
      • Teno J.M.
      • Roy J.
      • et al.
      Clinical and organizational factors associated with feeding tube use among nursing home residents with advanced cognitive impairment.
      In Hong Kong, concerns about feasibility and legal liability among hospital and residential care home providers to continue oral feeding to patients with high aspiration risk and/or progressive weight loss contribute to the persistence of widespread feeding tube usage.
      • Luk J.K.
      • Chan F.H.
      • Hui E.
      • Tse C.Y.
      The feeding paradox in advanced dementia: A local perspective.
      This study helps to strengthen the evidence for careful hand feeding, particularly for places where NGF is the predominant mode of long-term enteral feeding in dementia patients. Furthermore, it provides evidence to support CHF practices in hospitals and residential care homes not only for quality of life considerations but also demonstrated feasibility and the potential to reduce pneumonia risk in this population.

      Acknowledgments

      We thank the speech therapists who facilitated data collection for this study.

      Supplementary Data

      Supplementary Table 1Risk Factors for Survival: Univariate and Multivariate Analysis
      VariableUnivariate AnalysisMultivariate Analysis
      P ValueHR (95% CI)P ValueHR (95% CI)
      Age<.0011.03 (1.02, 1.05)<.0011.04 (1.03, 1.06)
      Male sex<.0011.50 (1.25, 1.79).071.25 (0.98, 1.59)
      Marital status (Ref: Married).02.01
       Single<.0010.49 (0.30, 0.78)<.0010.43 (0.26, 0.73)
       Divorced or separated.590.84 (0.46, 1.55).721.13 (0.59, 2.14)
       Widowed.060.83 (0.69, 1.01).140.83 (0.64, 1.07)
      Education (Ref: Illiterate).09.17
       Primary.031.26 (1.03, 1.55).051.25 (1.00, 1.56)
       Secondary.341.17 (0.85, 1.60).251.22 (0.87, 1.71)
       Tertiary.111.43 (0.92, 2.20).171.39 (0.87, 2.20)
      Living at home.510.93 (0.76, 1.15)
      Nasogastric tube feeding.551.06 (0.88, 1.27).181.15 (0.94, 1.39)
      Feeding problem (Ref: Behavioral).11
       Dysphagia.301.11 (0.91, 1.36)
       Both behavioral and dysphagia.240.85 (0.64, 1.12)
      Severity of dysphagia (Ref: Mild).66
       Mild-moderate.171.35 (0.88, 2.08)
       Moderate.341.17 (0.85, 1.63)
       Moderate-severe.271.22 (0.86, 1.75)
       Severe.191.23 (0.90, 1.68)
      Aspiration pneumonia<.0011.50 (1.25, 1.80)<.0011.39 (1.14, 1.69)
      Body mass index.010.97 (0.94, 0.99)
      Albumin level<.0010.94 (0.92, 0.95)<.0010.95 (0.94, 0.97)
      Lymphocyte count.110.88 (0.76, 1.03)
      Active pressure injury<.0011.36 (1.14, 1.63).011.29 (1.07, 1.56)
      History of stroke.760.97 (0.80, 1.18)
      Charlson comorbidity index<.0011.18 (1.09, 1.27)<.0011.19 (1.09, 1.29)
      FAST 7D-F (Ref: FAST 7A-C).051.23 (1.00, 1.51).081.21 (0.98, 1.50)
      ADL dependent.761.06 (0.75, 1.49)
      Advance care planning.051.46 (0.99, 2.13).411.51 (0.57, 4.00)
      Nonhospitalized DNACPR order.061.39 (0.99, 1.97).820.90 (0.37, 2.20)
      ADL, activities of daily living; DNACPR, do not attempt cardiopulmonary resuscitation; FAST, Functional Assessment Staging Tool; HR, hazard ratio; Ref, referent.
      Supplementary Table 2Risk Factors for Pneumonia: Univariate and Multivariate Analysis
      VariableUnivariate AnalysisMultivariate Analysis
      P ValueHR (95% CI)P ValueHR (95% CI)
      Age.951.00 (0.98, 1.02)
      Male sex<.0011.69 (1.35, 2.12).031.41 (1.04, 1.91)
      Marital status (Ref: Married).02.71
       Single.360.80 (0.50, 1.28).280.75 (0.44, 1.27)
       Divorced or separated.440.74 (0.34, 1.59).560.78 (0.33, 1.83)
       Widowed<.0010.69 (0.54, 0.87).580.92 (0.67, 1.25)
      Education (Ref: Illiterate).03.12
       Primary.791.04 (0.80, 1.35).460.89 (0.66, 1.20)
       Secondary.271.24 (0.85, 1.80).891.03 (0.69, 1.55)
       Tertiary.012.06 (1.25, 3.39).041.76 (1.03, 3.21)
      Living at home.020.67 (0.47, 0.94).150.76 (0.52, 1.10)
      Nasogastric tube feeding<.0011.47 (1.16, 1.86).011.41 (1.08, 1.85)
      Feeding problem (Ref: Behavioral).02.36
       Dysphagia.011.43 (1.10, 1.87).311.18 (0.86, 1.63)
       Both behavioral and dysphagia.051.39 (1.00, 1.93).161.31 (0.90, 1.89)
      Severity of dysphagia (Ref: Mild).53
       Mild-moderate.950.98 (0.55, 1.74)
       Moderate.171.32 (0.88, 1.99)
       Moderate-severe.221.32 (0.85, 2.03)
       Severe.441.17 (0.79, 1.72)
      Aspiration pneumonia<.0011.50 (1.20, 1.88).041.31 (1.01, 1.70)
      Body mass index.010.96 (0.93, 0.99).040.97 (0.93, 1.00)
      Albumin level.010.97 (0.95, 0.99).230.99 (0.97, 1.01)
      Lymphocyte count.620.97 (0.85, 1.10)
      Active pressure injury.251.14 (0.91, 1.43)
      History of stroke.221.16 (0.92, 1.47)
      Charlson comorbidity index.420.96 (0.86, 1.06)
      FAST 7D-F (Ref: FAST 7A-C).641.06 (0.83, 1.36)
      ADL dependent.351.25 (0.79, 1.99)
      Advance care planning.350.75 (0.41, 1.37)
      Nonhospitalized DNACPR order.761.08 (0.67, 1.74)
      ADL, activities of daily living; DNACPR, do not attempt cardiopulmonary resuscitation; FAST, Functional Assessment Staging Tool; HR, hazard ratio; Ref, referent.
      Supplementary Table 3Risk Factors for Pneumonia: Univariate and Multivariate Analysis (Removed Nasogastric Tube Feeding)
      VariableUnivariate AnalysisMultivariate Analysis
      P ValueHR (95% CI)P ValueHR (95% CI)
      Age.951.00 (0.98, 1.02)
      Male sex<.0011.69 (1.35, 2.12).081.32 (0.97, 1.79)
      Marital status (Ref: Married).02.67
       Single.360.80 (0.50, 1.28).730.91 (0.55, 1.52)
       Divorced or separated.440.74 (0.34, 1.59).311.18 (0.86, 1.64)
       Widowed<.0010.69 (0.54, 0.87).780.89 (0.38, 2.05)
      Education (Ref: Illiterate).03.16
       Primary.791.04 (0.80, 1.35).710.94 (0.70, 1.27)
       Secondary.271.24 (0.85, 1.80).751.07 (0.70, 1.63)
       Tertiary.012.06 (1.25, 3.39).041.77 (1.03, 3.03)
      Living at home.020.67 (0.47, 0.94).060.70 (0.48, 1.01)
      Feeding problem (Ref: Behavioral).02.08
       Dysphagia.011.43 (1.10, 1.87).061.34 (.98, 1.82)
       Both behavioral and dysphagia.051.39 (1.00, 1.93).041.48 (1.24, 2.13)
      Severity of dysphagia (Ref: Mild).53
       Mild-moderate.950.98 (0.55, 1.74)
       Moderate.171.32 (0.88, 1.99)
       Moderate-severe.221.32 (0.85, 2.03)
       Severe.441.17 (0.79, 1.72)
      Aspiration pneumonia<.0011.50 (1.20, 1.88).041.31 (1.01, 1.70)
      Body mass index.010.96 (0.93, 0.99).020.96 (0.93, 1.00)
      Albumin level.010.97 (0.95, 0.99).310.99 (0.97, 1.01)
      Lymphocyte count.620.97 (0.85, 1.10)
      Active pressure injury.251.14 (0.91, 1.43)
      History of stroke.221.16 (0.92, 1.47)
      Charlson comorbidity index.420.96 (0.86, 1.06)
      FAST 7D-F (Ref: FAST 7A-C).641.06 (0.83, 1.36)
      ADL dependent.351.25 (0.79, 1.99)
      Advance care planning.350.75 (0.41, 1.37)
      Nonhospitalized DNACPR order.761.08 (0.67, 1.74)
      ADL, activities of daily living; DNACPR, do not attempt cardiopulmonary resuscitation; FAST, Functional Assessment Staging Tool; HR, hazard ratio; Ref, referent.
      Supplementary Table 4Risk Factors for Pneumonia: Univariate and Multivariate Analysis (Removed Aspiration Pneumonia)
      VariableUnivariate AnalysisMultivariate Analysis
      P ValueHR (95% CI)P ValueHR (95% CI)
      Age.951.00 (0.98, 1.02)
      Male sex<.0011.69 (1.35, 2.12).021.44 (1.06, 1.95)
      Marital status (Ref: Married).02.62
       Single.360.80 (0.50, 1.28).560.86 (0.51, 1.43)
       Divorced or separated.440.74 (0.34, 1.59).401.15 (0.83, 1.59)
       Widowed<.0010.69 (0.54, 0.87).590.79 (0.34, 1.84)
      Education (Ref: Illiterate).03.17
       Primary.791.04 (0.80, 1.35).560.92 (0.68, 1.23)
       Secondary.271.24 (0.85, 1.80).661.10 (0.72, 1.68)
       Tertiary.012.06 (1.25, 3.39).051.71 (1.00, 2.92)
      Living at home.020.67 (0.47, 0.94).100.73 (0.50, 1.06)
      Nasogastric tube feeding<.0011.47 (1.16, 1.86).011.41 (1.08, 1.85)
      Feeding problem (Ref: Behavioral).02.21
       Dysphagia.011.43 (1.10, 1.87).121.28 (0.94, 1.75)
       Both behavioral and dysphagia.051.39 (1.00, 1.93).111.35 (0.93, 1.97)
      Severity of dysphagia (Ref: Mild).53
       Mild-moderate.950.98 (0.55, 1.74)
       Moderate.171.32 (0.88, 1.99)
       Moderate-severe.221.32 (0.85, 2.03)
       Severe.441.17 (0.79, 1.72)
      Body mass index.010.96 (0.93, 0.99).050.97 (0.93, 1.00)
      Albumin level.010.97 (0.95, 0.99).180.99 (0.96, 1.01)
      Lymphocyte count.620.97 (0.85, 1.10)
      Active pressure injury.251.14 (0.91, 1.43)
      History of stroke.221.16 (0.92, 1.47)
      Charlson comorbidity index.420.96 (0.86, 1.06)
      FAST 7D-F (Ref: FAST 7A-C).641.06 (0.83, 1.36)
      ADL dependent.351.25 (0.79, 1.99)
      Advance care planning.350.75 (0.41, 1.37)
      Nonhospitalized DNACPR order.761.08 (0.67, 1.74)
      ADL, activities of daily living; DNACPR, do not attempt cardiopulmonary resuscitation; FAST, Functional Assessment Staging Tool; HR, hazard ratio; Ref, referent.

      References

        • Mitchell S.L.
        • Teno J.M.
        • Kiely D.K.
        • et al.
        The clinical course of advanced dementia.
        N Eng J Med. 2009; 361: 1529-1538
        • Li I.
        Feeding tubes in patients with severe dementia.
        Am Fam Physician. 2002; 65: 1605-1611
        • Mitchell S.L.
        • Lawson F.M.
        Decision-making for long-term tube-feeding in cognitively impaired elderly people.
        Can Med Assoc J. 1999; 160: 1705-1709
        • Carey T.S.
        • Hanson L.
        • Garrett J.M.
        • et al.
        Expectations and outcomes of gastric feeding tubes.
        Am J Med. 2006; 119 (e11–e16): 527
        • Hanson L.C.
        • Garrett J.M.
        • Lewis C.
        • et al.
        Physicians' expectations of benefit from tube feeding.
        J Palliat Med. 2008; 11: 1130-1134
        • Finucane T.E.
        • Christmas C.
        • Travis K.
        Tube feeding in patients with advanced dementia: A review of the evidence.
        JAMA. 1999; 282: 1365-1370
        • Sampson E.L.
        • Candy B.
        • Jones L.
        Enteral tube feeding for older people with advanced dementia.
        Cochrane Database Syst Rev. 2009; 2: CD007209
        • Lee Y.F.
        • Hsu T.W.
        • Liang C.S.
        • et al.
        The efficacy and safety of tube feeding in advanced dementia patients: A systemic review and meta-analysis study.
        J Am Med Dir Assoc. 2021; 22: 357-363
        • American Geriatrics Society Ethics Committee
        • Clinical Practice and Models of Care Committee
        American Geriatrics Society feeding tubes in advanced dementia position statement.
        J Am Geriatr Soc. 2014; 62: 1590-1593
        • Volkert D.
        • Chourdakis M.
        • Faxen-Irving G.
        • et al.
        ESPEN guidelines on nutrition in dementia.
        Clin Nutr. 2015; 34: 1052-1073
        • Allan R.
        Oral feeding difficulties and dilemmas–a working party report.
        Clin Med (Lond). 2010; 10: 3
        • Australian and New Zealand Society for Geriatric Medicine
        • Australian and New Zealand Society for Geriatric Medicine
        Position statement–dysphagia and aspiration in older people.
        Australas J Ageing. 2011; 30: 98-103
        • Delegge M.H.
        Percutaneous endoscopic gastrostomy in the dementia patient: helpful or hindering?.
        Am J Gastroenterol. 2008; 103: 1018-1020
        • Lynch M.C.
        Is tube feeding futile in advanced dementia?.
        Linacre Q. 2016; 83: 283-307
        • Davies N.
        • Barrado-Martín Y.
        • Vickerstaff V.
        • et al.
        Enteral tube feeding for people with severe dementia.
        Cochrane Database Syst Rev. 2021; 8: CD013503
        • Alvarez-Fernández B.
        • García-Ordoñez M.A.
        • Martínez-Manzanares C.
        • Gómez-Huelgas R.
        Survival of a cohort of elderly patients with advanced dementia: Nasogastric tube feeding as a risk factor for mortality.
        Int J Geriatr Psychiatry. 2005; 20: 363-370
        • Chou H.H.
        • Tsou M.T.
        • Hwang L.C.
        Nasogastric tube feeding versus assisted hand feeding in-home healthcare older adults with severe dementia in Taiwan: A prognosis comparison.
        BMC Geriatr. 2020; 20: 60
        • Zhang Y.
        • Ma C.
        • Li C.
        • Chen Q.
        • Shen M.
        • Wang Y.
        Clinician's attitude to enteral nutrition with percutaneous endoscopic gastrostomy: A survey in China.
        J Health Popul Nutr. 2021; 40: 42
      1. Nasogastric tube feeding prevalence in Taiwan. National Health Insurance Administration. Ministry of Health and Welfare in Taiwan, 2017
        • Wong A.
        • Sowa P.M.
        • Banks M.D.
        • Bauer J.D.
        Home enteral nutrition in Singapore's long-term care homes—incidence, prevalence, cost, and staffing.
        Nutrients. 2019; 11: 2492
        • Luk J.K.H.
        • Chan W.K.
        • Ng W.C.
        • et al.
        Mortality and health services utilization among older people with advanced cognitive impairment living in residential care homes.
        Hong Kong Med J. 2013; 19: 518-524
      2. HA guidelines on life-sustaining treatment in the terminally ill 2015. Hospital Authority, 2015
        • Luk J.K.H.
        • Chan T.C.
        • Chan F.H.W.
        Careful hand feeding program in a geriatric step-down hospital in Hong Kong–is this feasible?.
        J Frailty Aging. 2021; 10: 303-304
        • Kuo S.
        • Rhodes R.L.
        • Mitchell S.L.
        • et al.
        Natural history of feeding-tube use in nursing home residents with advanced dementia.
        J Am Med Dir Assoc. 2009; 10: 264-270
        • Reisberg B.
        • Ferris S.H.
        • de Leon M.J.
        • Crook T.
        The global deterioration scale for assessment of primary degenerative dementia.
        Am J Psych. 1982; 139: 1136-1139
        • Volicer L.
        • Seltzer B.
        • Rheaume Y.
        • et al.
        Eating difficulties in patients with probable dementia of the Alzheimer type.
        J Geriatr Psychiatry Neurol. 1989; 2: 188-195
        • Stockdell R.
        • Amella E.J.
        The Edinburgh Feeding Evaluation in Dementia Scale: determining how much help people with dementia need at mealtime.
        Am J Nurs. 2008; 108: 46-54
        • Chan C.P.H.
        • Kwan Y.K.
        Feeding-swallowing issues in older adults with dementia.
        Asian J Gerontol Geriatr. 2014; 9: 80-84
        • Ward E.C.
        • Conroy A.L.
        Validity, reliability and responsivity of the Royal Brisbane Hospital outcome measure for swallowing.
        Asia Pac J Speech Lang Hear. 1999; 4: 109-129
        • Meier D.E.
        • Ahronheim J.C.
        • Morris J.
        • Baskin-Lyons S.
        • Morrison R.S.
        High short-term mortality in hospitalized patients with advanced dementia: Lack of benefit of tube feeding.
        Arch Intern Med. 2001; 161: 594-599
        • Teno J.M.
        • Gozalo P.L.
        • Mitchell S.L.
        • et al.
        Does feeding tube insertion and its timing improve survival?.
        J Am Geriatr Soc. 2012; 60: 1918-1921
        • Ticinesi A.
        • Nouvenne A.
        • Lauretani F.
        • et al.
        Survival in older adults with dementia and eating problems: To PEG or not to PEG?.
        Clin Nutr. 2016; 35: 1512-1516
        • Payne M.
        • Morley J.E.
        Editorial: Dysphagia, dementia and frailty.
        J Nutr Health Aging. 2018; 22: 562-565
        • Blumenstein I.
        • Shastri Y.M.
        • Stein J.
        Gastroenteric tube feeding: techniques, problems and solutions.
        World J Gastroenterol. 2014; 20: 8505-8524
        • Gomes G.F.
        • Pisani J.C.
        • Macedo E.D.
        • Campos A.C.
        The nasogastric feeding tube as a risk factor for aspiration and aspiration pneumonia.
        Curr Opin Clin Nutr Metab Care. 2003; 6: 327-333
        • Mitchell S.L.
        • Mor V.
        • Gozalo P.L.
        • et al.
        Tube feeding in US nursing home residents with advanced dementia, 2000–2014.
        JAMA. 2016; 316: 769-770
        • Ojo O.
        The challenges of home enteral tube feeding: A global perspective.
        Nutrients. 2015; 7: 2524-2538
        • Best C.
        • Hitchings H.
        Enteral tube feeding–from hospital to home.
        Br J Nurs. 2010; 19: 174
        • Mitchell S.L.
        • Teno J.M.
        • Roy J.
        • et al.
        Clinical and organizational factors associated with feeding tube use among nursing home residents with advanced cognitive impairment.
        JAMA. 2003; 290: 73-80
        • Luk J.K.
        • Chan F.H.
        • Hui E.
        • Tse C.Y.
        The feeding paradox in advanced dementia: A local perspective.
        Hong Kong Med J. 2017; 23: 306-310