Predictive factors associated with oral intake ability in gastrostomy patients under long-term care

Abstract

Objective

To determine the physical indicators associated with oral intake status and swallowing function in gastrostomy patients under long-term care.

Design

Cross-sectional study.

Setting

Thirty-one hospitals that perform gastrostomy insertion, replacement and management.

Participants

A total of 117 respondents from 31 hospitals in Japan underwent gastrostomy tube replacement and management between September 2012 and January 2014. Each participant underwent a gastrostomy at least 6 months prior to the study, and received long-term care either at home, a care facility, or a hospital.

Measurements

We conducted a questionnaire survey at Japanese hospitals and used the data obtained from 117 respondents for analysis. The survey was conducted using a questionnaire form that collected information about the following items: oral intake status, sex, age, disease history, number of days elapsed since gastrostomy, residence status, modified Rankin Scale score, consciousness, oral hygiene status, articulation and phonation, voluntary saliva swallow, Modified Water Swallow Test, and Food Test.

Results

Results revealed significant differences in modified Rankin Scale scores, sputum production, articulation and phonation, and voluntary saliva swallowing between patients who were orally fed and those who were not. Moreover, sputum production and voluntary saliva swallowing were strongly associated with oral intake status. Finally, sputum production, articulation and phonation, and voluntary saliva swallowing were strongly associated with swallowing function test results.

Conclusion

Results from this study suggested that sputum production, articulation and phonation, and voluntary saliva swallowing could be used as indicators for estimating oral intake status and swallowing function in gastrostomy patients under long-term care.

Introduction

Since its first report in 1980 (1), percutaneous endoscopic gastrostomy (PEG) has been performed worldwide as a stable long-term tube feeding method in patients with difficulties in oral nutrition intake because of severe disorders. In some countries, the number of gastrostomy cases has continuously increased with the rise in the older population. In 2001, over 216,000 PEGs were performed in the United States (2). In 2014, the estimated number was 96,000–119,000 in Japan (3). This increase has also resulted in an increased proportion of tube-fed patients in hospitals and facilities, with reported rates (per 100 beds) of 11.6 in nursing homes, 7.4 in long-term care facilities, 36.3 in sanatorium medical facilities, and 7.9 in rehabilitation facilities (4).

Gastrostomy is considered less inconvenient, less uncomfortable, and interferes less with family life and social activities (5, 6) compared with nasogastric tube-assisted feeding (i.e., a method of providing enteral nutrition via nasogastric tubing). For these reasons, gastrostomy could potentially improve the level of independent living (6). Conversely, there is supposedly no significant difference in mortality or the incidence of aspiration pneumonia between gastrostomy feeding and nasogastric tube-assisted feeding (5). Therefore, in recent years, gastrostomy has been considered according to the following guidelines: “each case should be considered on its own merits, taking into account the clinical situation, diagnosis, prognosis, ethical issues, the expected effect on the patient's quality of life (QOL) and the patient's own wishes (7).”

A study that examined the QOL of patients who had undergone gastrostomy (hereafter, gastrostomy patients) suggested that oral intake is strongly associated with improved QOL [odds ratio (OR): 4.285] and ‘enjoying one's own life' [OR: 4.171] (8). That is, whether or not oral intake can be continued greatly influences life after gastrostomy for these patients. However, in Japan, many patients with gastrostomy under long-term care continue to follow physician's instructions to ingest something or nothing orally, which was determined based on swallowing function at the time of discharge from the hospital. However, patients who do not orally ingest after gastrostomy rarely have the opportunity to undergo swallowing function tests, including the videofluorographic swallowing study (VFSS) and fiberoptic endoscopic evaluation of swallowing (FEES), as well as a simplified swallowing screening test (4). In fact, the hospitals generally do not inspect swallowing function or verify the validity of oral intake restriction prior to gastrostomy tube replacement unless there are exceptional circumstances, such as strong demand from patient's family. This could be related to issues such as cost and labor, as well as aspiration risk at time of examination. However, because swallowing function is influenced by various factors including nutritional status (9, 10), respiratory conditions (11), number of days from onset (12, 13), and medication status (14, 15), the suitability of oral intake should be discussed in each case according to changes in general patient condition as well as their environment.

If some physical findings suggest that a patient might be able to regain oral intake, the patient should be given the opportunity to undergo swallowing function tests so that he or she would no longer have to avoid oral intake. Conversely, if physical findings suggest that continued oral intake for a patient who has been orally fed is unfeasible, the suitability of oral intake should be reconsidered by performing swallowing function tests to avoid the risk of developing aspiration pneumonia.

The goal of this study was to present physical findings that may serve as indicators for whether or not a gastrostomy patient under long-term care is capable of oral intake. To this end, we conducted a multi-center survey targeting these patients. By comparing orally fed patients with those who are not, we determined the factors that are strongly associated with oral intake status. In addition, we applied several swallowing screening tests at the timing of replacement of gastrostomy as research purpose. Then, we also examined several factors strongly associated with swallowing screening test results to determine the functional factors that might prevent oral intake.

Methods

Participants

We contacted 31 hospitals in Japan that performed gastrostomy insertion, replacement and management between September 2012 and January 2014 and requested their participation in a questionnaire survey regarding the physical findings of gastrostomy patients. Each patient had undergone a gastrostomy at least 6 months prior to the study period and received long-term care either at home, a care facility, or hospital. General patient status was evaluated by medical staff during hospitalization for gastrostomy tube replacement.

Assessment and test methods

The survey was conducted using a questionnaire form that collected information on the following items: oral intake status (whether the participant was orally ingesting anything on a routine basis), sex, age, disease history (stroke, dementia, aspiration pneumonia, Parkinson's disease), number of days elapsed since gastrostomy (days), residence status (home, care facility, or hospital), modified Rankin Scale (mRS) score ranging from independent to fully dependent (0 = fully independent; 1 = minor symptoms, not affecting lifestyle; 2 = minor handicap, but independent in self-care; 3 = moderate handicap, needing some help with activities of daily living (ADLs); 4 = needing much help with ADLs; 5 = needing constant attention day and night) (16, 17), consciousness (the Japan Coma Scale (JCS); 0 = alert and conscious; 1 = awake status without stimulation; 2 = awake status when stimulated but falls asleep with no stimulation; 3 = asleep status even with stimulation), oral hygiene status (0 = good, 1 = poor), sputum production (0 = none; 1 = small amount; 2 = large amount), articulation and phonation (0 = non-disability; 1 = dysarthria and/or hoarseness; 2 = phonation disability), swallowing within 30 s when instructed to voluntarily swallow saliva (voluntary saliva swallow) (0 = incapable; 1 = capable). These questionnaires were filled out by the attending physician or nurse.

In addition, they performed the Modified Water Swallow Test (MWST) and Food Test (FT) (18) with the safety of the participant taken into consideration, regardless of oral intake status. The MWST, which involves squirting 3 cc of water into the floor of the oral cavity and assessing swallowing condition, is scored as follows: 1 = cannot swallow at all; 2 = dyspnea (difficulty breathing) after swallowing; 3 = coughing or wet hoarseness after swallowing; 4 = can swallow with no abnormality after swallowing; 5 = can swallow with no abnormality after swallowing, and can also voluntarily swallow saliva twice within 30 s afterward. The FT assesses swallowing condition by placing approximately one teaspoon of pudding on the tongue, and is scored as follows: 1 = cannot swallow at all; 2 = dyspnea (difficulty breathing) after swallowing; 3 = coughing or wet hoarseness, or 25% or more pudding is retained in the mouth after swallowing; 4 = can swallow with no abnormality after swallowing; 5 = can swallow with no abnormality after swallowing, and can also swallow saliva voluntarily twice within 30 s afterward.

Statistical analysis

Patients were divided into two groups according to oral intake status, and data were compared using the unpaired t-test or chi-square test. Furthermore, logistic regression analysis was performed to examine factors associated with oral intake status. For analysis of factors associated with MWST and FT scores, multiple regression analysis was performed. p < 0.05 was considered statistically significant. SPSS Statistics 22 Software (IBM Corporation, Armonk, NY, USA) was used for statistical processing.

Ethical considerations

This study approval was performed at the Nihon University School of Dentistry (Approval no.: 2011-4). We obtained written and oral consent from all participants and their families.

Results

We collected questionnaire forms from a total of 302 respondents. After excluding those with blank survey items, we incorporated forms from 117 respondents (31 men, 86 women) into our analysis. The participant age ranged from 58 to 101 years (mean ± SD, 83.9 ± 9.5 years), and the number of days elapsed since gastrostomy ranged from 180 to 3360 days (mean ± SD, 975.0 ± 736.6 days). The primary disease indications for gastrostomy insertion were cerebral infarction (n = 38), cerebral hemorrhage (n = 15), subarachnoid hemorrhage (n = 2), dementia (n = 31), Parkinson's disease (n = 12), aspiration pneumonia (n = 9), and other causes (n = 10). Forty-three participants were classified as the oral intake group; 5 of these orally ingested nutrition more than from the gastrostoma, and 38 participants primarily obtained nutrition from then gastrostoma, with a small amount of food ingested for comfort or swallowing training. A total of 74 participants were classified as the no oral intake group because they ingested nutrition only from the gastrostomy.

Comparing the two groups according to oral intake status revealed significant differences in the history of dementia, mRS, sputum production, articulation and phonation, voluntary saliva swallowing, MWST, and FT (Table 1).

Table 1.

Characteristics and Comparison of the two groups by oral intake status

		ALL (n = 117)	oral intake (n = 43)	no oral intake (n = 74)	p-value
Sex	male / female, n	31 / 86	13 / 30	18 / 56	.485
Age	mean ± SD, years	83.9 ± 9.5	83.8 ± 9.4	84.0 ± 9.7	.769
Elapsed days from gastrostomy	mean ± SD, days	975.0 ± 736.6	751.1 ± 674.9	1089.1 ± 746.5	.051
History of disease	stroke, %	47.9	69.8	60.8	.181
	dementia, %	54.7	44.2	60.8	.002*
	aspiration pneumonia, %	34.2	30.2	36.5	.345
	Parkinson’s disease, %	10.3	7.0	12.2	.212
Residence status	home / care facility / hospital, %	28.2 / 45.3 / 26.5	33.3 / 50.0 / 16.7	26.2 / 43.1 / 30.8	.063
mRS	0 / 1 / 2 / 3 / 4 / 5, %	0.9 / 0.9 / 0.0 / 3.4 / 78.6	2.3 / 0.0 / 0.0 / 4.7 / 30.2 / 62.8	0.0 / 1.4 / 0.0 / 2.7 / 8.1 / 87.8	<.001*
JCS	0 / 1 / 2 / 3, %	30.8 / 39.3 / 20.5 / 9.4	39.5 / 44.2 / 11.6 / 46.5	25.7 / 36.5 / 25.7 / 12.2	.096
Oral hygiene status	0 / 1, %	54.9 / 45.1	55.8 / 44.2	54.1 / 45.9	.809
Sputum production	0 / 1 / 2, %	30.0 / 47.9 / 22.2	39.5 / 53.5 / 7.0	24.3 / 44.6 / 31.1	<.001*
Articulation and Phonation	0 / 1 / 2, %	18.8 / 37.6 / 43.6	30.2 / 46.5 / 23.3	12.2 / 32.4 / 55.4	<.001*
Saliva swallow	0 / 1, %	20.9 / 79.1	34.9 / 65.1	6.8 / 93.2	<.001*
MWST	1 / 2 / 3 / 4 / 5, %	17.9 / 3.4 / 26.5 / 13.7 / 7.7	7.7 / 2.6 / 38.4 / 28.2 / 23.1	42.9 / 7.1 / 38.1 / 11.9 / 0.0	<.001*
	unexecuted, n	36	1	35
FT	1 / 2 / 3 / 4 / 5, %	17.1 / 0.9 / 9.4 / 22.2 / 6.8	14.0/ 2.3 / 14.0 / 51.2 / 18.6	60.9 / 0.0 / 21.7/ 17.4 / 0.0	<.001*
	unexecuted, n	51	0	51

mRS: modified Rankin Scale score, JCS: Japan Coma Scale, MWST: Modified Water Swallow Test, FT: Food Test; unexecuted; Cases which the attending physician or nurse refrained from the swallowing test due to their conditions; * chi-square test, p <.05

Factors associated with oral intake status were examined using physical findings, revealing significant differences in the two-group comparison as explanatory variables. As a result, significant associations were found with sputum production (OR [95% confidence interval (CI)]: 0.442 [0.231–0.847]) and voluntary saliva swallow (4.833 [1.429–16.350]) (Table 2). Moreover, in the multiple regression analysis using the MWST score as an objective variable, articulation and phonation [standard partial regression coefficient (β) (95% CI): −0.273 (−0.840 to −0.076)] and voluntary saliva swallowing [0.357 (0.387–1.584)] had a significant association. Conversely, the multiple regression analysis, in which the FT score served as the objective variable, showed that only sputum production [−0.313 (−1.182 to −0.157)] was significantly different (Tables 3, 4).

Table 2.

Logistic regression analysis for oral intake status

n=117
	B	SE	OR (95%Cl)	p-value
mRS	-.040	.300	.961 (.534 – 1.73)	.895
Sputum production	-.816	.332	.442 (.231 – .847)	.014*
Articulation and phonation	-.559	.305	.572 (.315 – 1.04)	.067
Saliva swallow	1.576	.622	4.833 (1.429 – 16.35)	.011*

^*p <.05

Table 3.

Multiple regression analysis for MWST

n=81
	B	SE	β (95%Cl)	p-value
mRS	-.008	.135	-.006 (-.276 – .261)	.954
Sputum production	-.323	.176	-.178 (-.673 – .026)	.069
Articulation and phonation	-.458	.192	-.273 (-.840 – -.076)	.019*
Saliva swallow	.986	.301	.357 (.387 – 1.584)	.002*

^*p <.05

Table 4.

Multiple regression analysis for FT

n=66
	B	SE	β (95%Cl)	p-value
mRS	.044	.227	.025 (-.411 – .498)	.849
Sputum production	-.670	.256	-.313 (-1.182– -.157)	.011*
Articulation and phonation	-.356	.253	-.184 (-.860 – .149)	.164
Saliva swallow	.622	.408	.208 (-.193 – 1.437)	.132

^*p <.05

Discussion

Orally fed patients exhibited better outcomes with regard to mRS, sputum production, articulation and phonation, and voluntary saliva swallowing. In particular, results suggested that sputum production and voluntary saliva swallow were strongly associated with oral intake status. Moreover, MWST scores were significantly associated with voluntary saliva swallowing movement and articulation/phonation, whereas FT scores were significantly associated with sputum production.

Oral intake status in gastrostomy patients under long-term care

Previous studies have reported that younger age (19, 20, 21, 22) and daily living abilities (19) are positive factors for regaining oral intake after gastrostomy placement. In the present study, however, age did not significantly differ with oral intake status. This might be because our study participants were older than those in previous studies. The fact that most long-term residents at care facilities or hospitals in Japan were geriatric patients likely affected the age composition of our study participants. Although differences were observed in mRS between the two groups, the multiple comparison analysis did not reveal a significant association. Therefore, our results suggested that age and daily living abilities were not strong factors affecting oral intake status in older gastrostomy patients under long-term care.

To initiate voluntary saliva swallowing as instructed, patients require proper activity of the cerebral cortex to understand the instruction, as well as proper functioning of the deglutition center and swallowing-associated muscles to execute the given commands. Accordingly, it is possible to verify activities of these organs by assessing voluntary saliva swallowing movement. We surmise that many of the orally fed patients maintained normal activity in these organs, allowing them to pass this test. However, because voluntary saliva swallow is affected by cognitive function (i.e., understanding verbal instructions), as well as by the amount of saliva in the oral cavity (23), difficulty in voluntarily swallowing saliva does not necessarily indicate an inability to perform oral intake, which was consistent with some of the orally fed patients who could not pass the test. Conversely, it is possible that some patients were not allowed oral intake because of a delayed swallowing reflex or insufficient swallowing pressure, despite the ability to pass the test. Therefore, it is difficult to determine whether to permit oral intake solely based on the patient's ability to voluntarily swallow saliva. Nevertheless, in the assessment of swallowing function in individuals who have not yet regained oral intake, voluntary saliva swallowing remains a clinically useful measure to confirm performance of organs related to swallowing without any risks.

Clinically, increased sputum production is a typical symptom of pneumonia. However, there were no patients in the present study with respiratory diseases. Therefore, it is possible that microaspiration (i.e., when unswallowed saliva flows into the trachea) (24) was the cause of sputum production. Microaspiration is often observed in severe stroke patients with swallowing disorders and patients with advanced Parkinson's disease or dementia (25), and is reportedly associated with a high risk of aspiration (26). For this reason, it is likely that many of our patients with increased sputum production had severe physical dysfunctions (including swallowing function disorders) and were, thus, incapable of oral intake.

There were significant differences in articulation and phonation in the between-group comparison with regard to oral intake status. A previous study reported that articulation and phonation disorders are associated with swallowing function and aspiration risk (27). However, our logistic regression analysis results suggested that, compared with articulation and phonation, voluntary saliva swallowing and sputum production more accurately reflected the characteristics of gastrostomy patients with regard to oral intake. Therefore, they could be used as indicators for whether a gastrostomy patient under longterm care has the ability to perform oral intake.

The association between swallowing function tests and physical findings

Both the MWST and FT are screening tests used to assess swallowing function (18). While neither the MWST nor FT is superior to the other in terms of suitability to diagnose dysphagia, the ability that is assessed at the time of testing differs. The MWST requires the patient to let water, which is injected into the vestibule of the mouth, flow through the pharynx, as well as a quick swallowing reflex to manage the water that flows into the pharynx. The FT requires the patient to transport a bolus of solid food on the tongue into the pharynx, as well as swallowing pressure to push the bolus into the esophagus. If the patient fails one of these tests, even if he or she is able to pass the other, oral intake can be initiated by modifying food consistency or contriving food form. Therefore, it is considered desirable to perform both tests.

The performance rates for the MWST and FT were low among patients who were not orally fed. In a previous study, the performance rates of swallowing function tests in tube-fed patients in long-term care facilities were also similarly low, which was consistent with the present study (4). Therefore, long-term patients who are prohibited from oral feeding after gastrostomy have limited opportunities for oral intake assessment, and medical workers are hesitant to perform such assessments because of the fear of aspiration occurring during examination. However, if the criteria for relevant decision making are not clearly established and are based on the experience and intuition of medical workers, then these decisions could inevitably and unnecessarily prevent gastrostomy patients from regaining their ability to lead a fulfilling life. Results from the present study suggested that voluntary saliva swallowing, articulation and phonation, and sputum production could serve as effective indicators for assessing swallowing function of gastrostomy patients under long-term care. Although associations between these factors and swallowing function have been previously reported, their associations with a swallowing screening test have not been previously reported.

The MWST requires a series of movements similar to those involved in voluntary saliva swallowing, such as gathering water injected into the vestibule of the mouth onto the tongue, sending it down to the pharynx, and swallowing. Furthermore, sensory functions of the pharynx and larynx become critical for enabling the patient to swallow water that flows into his/her pharynx in a timely manner. However, patients with hoarseness cannot elicit reflex movements or have delayed reflexes owing to larynx paralysis. We presume that this was responsible for the significant associations between MWST scores and voluntary saliva swallowing movement, as well as articulation/ phonation. Moreover, there was a significant association between FT scores and sputum production, possibly because the orally fed patients had almost no sputum production. We surmise that the majority of patients were able to continue oral intake without aspiration of food, i.e., a possible cause of increased sputum. Therefore, many were also able to swallow pudding without aspiration during FT. Consequently, gastrostomy patients under long-term care might find it difficult to regain oral intake because of impaired functions, such as skilled tongue movement, quick swallowing reflex movements, and having sufficient force and a quick enough reaction to cough up aspiration products.

There are some limitations to this study. First, because swallowing function tests involving VFSS or FEES were not performed in the present study, individual swallowing abilities were not examined and analyzed in detail. Furthermore, because the present study employed a cross-sectional research design, the causal relationships were not determined. In fact, we could not determine whether some of the statistically significant characteristics observed in our patients contributed to oral intake ability/disability or an orally fed status. A follow-up study is necessary to clarify this point. Finally, the present study had a small sample size, the questionnaire consisted of a small number of items, and both the MWST and FT had low performance rates. Accordingly, we could not eliminate the influence of selection bias. Thus, we hope to perform a followup study with a larger sample size and an increased number of investigation items in the future.

Conclusion

The present study found a strong association between oral intake status and sputum production and voluntary saliva swallowing. Moreover, sputum production, voluntary saliva swallowing, and articulation and phonation were strongly associated with swallowing function test results, suggesting that these factors could be used as indicators for estimating oral intake status and swallowing function in gastrostomy patients under long-term care.

Sources of funding: Health and Labor Sciences Research Grant, Comprehensive Research on Aging and Health in 2011- 2013

Conflict of interest: none declare.