Synthesis of research on ENFit gastrostomy tubes with potential implications for US patients using these devices

Abstract

Misconnections between enteral devices and other medical devices have been associated with patient death and serious injuries. To minimize such misconnections, the design of connectors on enteral devices has been standardized. The most common adaptation of the standardized enteral connector is called ENFit. Gastrostomy tubes (G-tubes), which may or may not possess the ENFit connector, are increasingly used to deliver commercial and blenderized diets in home settings to enteral device users. To investigate and compare the performance of G-tubes with and without ENFit connectors, research investigations have recently been performed. However, synthesis of such investigations and quantitative discussion of the consequences of transitioning to ENFit-based G-tube devices has not yet occurred. Here we review the research findings from these studies, with data on patient practices from a Mayo Clinic survey, to estimate the impact on tube feeders in home settings of transitioning to ENFit-based G-tube devices. Extrapolating the findings from these studies to US enteral G-tube patients, 2.5%–8.6% of adult patients and 0.2%–1.9% of pediatric patients may experience perceptible slowing in their gravity feeds if using ENFit-based G-tube devices. About 2.5%–8.6% of adult patients and 0.5%–5.5% of pediatric patients (or their caregivers) may need to push with perceptibly more force for syringe push-based feeding using ENFit-based G-tube devices. Lastly, the article offers suggestions for patients and device manufacturers.

INTRODUCTION

Enteral nutrition entails delivering nutrients and fluid via a tube inserted into the gastrointestinal (GI) tract to treat patients who cannot meet their nutrition needs orally.¹ These patients often include those affected by neurologic or neuromuscular dysfunction; head, neck, or upper GI cancers; anorexia; food allergies; inflammatory bowel disease; and failure to thrive. A misconnection^2,3 is an inappropriate connection occurring between two connectors that attach devices with different intended uses or applications and can occur in enteral feeding systems and nonenteral systems (Figure 1). Such misconnections can result in patient injury or death.⁴ Multiple alerts from The Joint Commission, the Emergency Care Research Institute (ECRI), the US Food and Drug Administration (FDA), and the US Pharmacopeia have been issued to increase awareness around such misconnections.⁴ Such alerts led to the creation of an international consensus effort for creating an enteral connector design that would be incompatible with other connectors found in hospital settings.⁵ As part of that effort, the Association for the Advancement of Medical Instrumentation (AAMI) published a provisional standard under AAMI CN3 in 2014, which the FDA recognized. The establishment of the general requirements for small-bore connector standard International Organization for Standardization (ISO) 80369-1 ensued, followed by standard ISO 80369-3 that was specifically intended for enteral connectors.^6,7 The ISO 80369-3 standard was approved worldwide in 2015 and published in 2016. It was also recognized by the FDA in 2016. The FDA cleared the first devices that used the new ISO connector design for the US market through the 510(k) premarket notification process in 2014.⁸ Figure 2 shows the timeline of the introduction of the ISO standard along with salient events and research in the last decade pertaining to the ISO standard. To reduce enteral misconnections, it is expected that nearly 500,000 feeding tube users in the US will be transitioning to the new feeding tubes with the ISO 80369-3 connectors.^9,10 Although the ISO 80369-3 standard stipulates an enteral small-bore connector (internal) diameter of 2.9 mm, the standard does not specify all dimensions of a finished gastrostomy tube (G-tube) device (eg, the length of the connector outside the physical connection) nor the direction of male-female connections. ENFit, the most common adaptation of the ISO 80369-3 standard, has received considerable attention from stakeholders, including patients, primary caregivers, healthcare delivery organizations, patient advocates, and manufacturers.^6,7,11-20

FIGURE 1.

Graphics show three different misconnections of various medical devices with enteral devices (adapted from https://www.fda.gov/medical-devices/medical-device-connectors/examples-medical-device-misconnections). These examples are not all inclusive

FIGURE 2.

The timeline of seminal events or studies on ENFit connectors and devices. The inset is a graphical representation of the ENFit connector adapted with permission from the GEDSA webpage. ASPEN, American Society for Parenteral and Enteral Nutrition; ECRI, Emergency Care Research Institute; FDA, US Food and Drug Administration; GEDSA, Global Enteral Device Supplier Association; ISO, International Organization for Standardization; JC, Joint Commission.

https://stayconnected.org/wp-content/uploads/2016/10/GEDSAStayconnected-Checklist-1-page.pdf

THE CHANGING LANDSCAPE OF HOME ENTERAL NUTRITION (HEN)

Concerns have been raised by some stakeholders about the performance of the ENFit-based G-tube devices in delivering HEN.⁶ HEN has been increasingly used because it reduces the time spent in a hospital.¹⁰ HEN users have raised concerns that the new G-tube devices will result in increased time of feeding, increased likelihood of clogging, difficulty in pushing diets because of a higher resistance to flow, cleaning problems, and the potential for drug dosing inaccuracy because of ENFit’s smaller inner diameter and direction of flow.^4,13 To investigate these concerns, several research studies into the performance of ENFit-based G-tube devices have been executed over the last five years. The objective of this article is to review the existing body of literature in the context of the impact of this transition on US patients who use G-tube devices.

REVIEW METHODOLOGY

To review the existing literature, four resources were used: Google Scholar, PubMed, Web of Knowledge, and the Oley Foundation. Keywords used in electronic searches were “ENFit” and “ISO 80369.” Since the standard was not designated until after 2010, the review was limited to January 2010–December 2020. The exclusion criteria and total articles identified are provided in Figure 3.

FIGURE 3.

Flow diagram for identifying relevant articles that discuss enteral device performance in the context of the transition to ENFit. ISO, International Organization for Standardization

REVIEW OF LITERATURE

Figure 4 shows the most common designs of G-tube devices and the various transition scenarios. Table 1 summarizes the testing that has been performed with various ENFit-based G-tube devices, in chronological order. Most of the testing has been performed on G-tubes in the 14 French (Fr) to 24 Fr size range. Of the various modes of feeding, gravity and push mode have been extensively studied.

FIGURE 4.

(A) A commonly used legacy syringe transitioning to an ENFit-based syringe. The plungers for the syringes have not been shown. (B) A bolus legacy extension set for low-profile G-tubes transitioning to the ENFit-based connection set. Note the short length of the connector section. Also, the low-profile G-tubes themselves remain unaffected by the transition. (C) A percutaneous endoscopic gastrostomy tube with long or short transition sections transitioning to ENFit-based devices with a relatively long transition section (figure adapted from Guha et al, 2020²³). G-tube, gastrostomy tube

TABLE 1.

Summarizes the studies found in literature to date with legacy and ENFit-based G-tube devices that used commercial or home-based diets, excluding breast milk

Organization	Device type investigated	Device sizes	Type of testing	Type of diet	Publication reference
GEDSA	Standard G-tube devices, low-profile G-tube devices	18/20 Fr	Gravity, force/pressure	Commercial formulations Commercial blenderized diets	Guenter and McMichael20
Mayo Clinic	Standard G-tube devicesa	20 Fr	Gravity	Commercial formulations Commercial blenderized diets Mayo blenderized diet	Hurt et al5
Mayo Clinic	Standard G-tube devices,a low-profile G-tube devicesa	20 Fr, 24 Fr	Force/pressure	Commercial formulations Commercial blenderized diets Mayo blenderized diet	Mundi et al6
Mayo Clinic	Standard G-tube devices, low-profile G-tube devices	14 Fr, 18 Fr,b 20 Fr, 24 Fr	Gravity	Commercial formulations Commercial blenderized diets	Mundi et al24
FDA	Standard G-tube devices, low-profile G-tube devices	14 Fr, 18 Fr, 20 Fr, 24 Fr	Gravity, force/pressure	Commercial formulations Commercial blenderized diets	Guha et al19
FDA	Standard G-tube devices, low-profile G-tube devices	14 Fr, 18 Fr, 20 Fr, 24 Fr	Gravity	Commercial formulations Commercial blenderized diets	Guha et al22
Mayo Clinic	Standard G-tube devices, low-profile G-tube devices	14 Fr, 18 Fr, 20 Fr, 24 Fr	Force/pressure	Commercial formulations Commercial blenderized diets	Mundi et al21
Japan, new connector working group	Standard G-tube devices,a low-profile G-tube devicesa	20 Fr, 24 Fr	Force/pressure	Commercial semisolid formulations	Maruyama et al10
FDA	Standard G-tube devices, low-profile G-tube devices	14 Fr, 18 Fr, 20 Fr, 24 Fr	Gravity, force/pressure	Home-based blenderized diets	Guha et al12

Abbreviations: FDA, US Food and Drug Administration; Fr, French; GEDSA, Global Enteral Device Supplier Association; G-tube, gastrostomy tube.

^aIndicates studies that used prototype ENFit connector designs rather than final commercialized designs.

^bIndicates limited studies that excluded some device types.

The first study, performed by Global Enteral Device Supplier Association (GEDSA) and described in a non-peer reviewed newsletter, did not find any difference in the administration of commercial blenderized diets through ENFit-based G-tube devices and the older devices, henceforth referred to as “legacy” devices.²¹ The Mayo Clinic performed flow rate testing on various prototype devices with two ENFit connector designs⁶ and found that one design, possessing a longer connector section, was consistently slower compared with another ENFit-based G-tube device design and legacy G-tube device designs by as much as 270% in some cases. Another study by the Mayo Clinic, involving force testing with prototype devices,²² revealed that, for commercial liquid diets as well as commercial blenderized diets, the differences in force required between legacy and ENFit-based G-tube devices were minor (<3 N; ie, <1 pound force [lbf]), but for homemade blenderized diets the force required by ENFit-based G-tube devices was higher by 7–10 N, that is, >1.5 lbf. Both of these findings raised concerns that the migration to ENFit-based G-tube devices could impact the HEN patients, particularly those who had been using feeding tubes with a larger inner diameter.

The Mayo Clinic and the FDA then collaborated to create a general framework and protocol to be used by each organization to independently test a large variety of legacy and ENFit-based G-tube devices. A range of device types (eg, standard and low-profile G-tube devices), device sizes (eg, 14–24 Fr), diets, and feeding modes (gravity and push) were chosen in the protocol. Four independent studies (two each by the Mayo Clinic and the FDA) based upon this framework were subsequently performed. The results are summarized in Table 1. In subsequent discussion, and in figures such as Figure 4, legacy devices are represented by “L” and ENFit-based G-tube devices by “E.” Results showed that for gravity feeds, with one exception, legacy and ENFit-based G-tube devices exhibited similar flow rates in the 14–20 Fr device size range across 10 commercially available diets.²² One 14 Fr ENFit-based G-tube device was found to be three to four times slower than the corresponding legacy device. For the 24 Fr size, the results were more varied. Some ENFit-based G-tube devices had flow rates comparable to those of legacy devices, whereas others were about 50% slower. The FDA concluded that the decreased flow rate was exacerbated with thicker diets.²² Use of the same protocol at the Mayo Clinic and the FDA helped enable a meaningful comparison across laboratories.

For the same device size and diet, the flow rate at the Mayo Clinic was typically 20%–30% lower than that at the FDA (an example of 18 Fr Isosource 1.5 is shown in Figure 5A). The reason for the difference in the flow rates is likely the variability in the few lots of the commercial formulas used. This difference, along with the finding that testing with water does not constitute worst case flow rate testing, underscores the importance of developing standardized fluids for flow rate testing for G-tube devices. Other trends between the Mayo Clinic and the FDA were consistent. For example, both labs observed that for larger diameter devices, such as 24 Fr, flow rates through legacy G-tubes L1 and L2 were similar to those through ENFit-based G-tubes E1 and E2, and the flow rate through legacy G-tube device L3 was consistently the fastest across multiple diets and device sizes (Figure 5B).

FIGURE 5.

(A) Flow rate data from FDA and Mayo Clinic for Isosource 1.5 at 18 Fr. (B) Flow rate data from FDA data with Peptamen AF at two device sizes across legacy and ENFit devices. (C) Force data from FDA with two blenderized diets at two device sizes across legacy and ENFit devices. L1–L4, E1, and E2 were standard G-tubes and L5 and E3 were low-profile G-tubes. E, ENFit device; FDA, US Food and Drug Administration; Fr, French; G-tube, gastrostomy tube; L, legacy device

For push mode, the Mayo Clinic and the FDA had several differences in their protocols. The Mayo Clinic primarily focused on commercial diets, whereas the FDA obtained recipes used by patients for their home-based diets and recreated them in a laboratory setting using different blenders.^12,22 Additionally, unlike gravity testing, in which the protocols were similar across both labs, equipment differences did not enable the two groups to use similar protocols for force testing. The Mayo Clinic used a force testing machine that could not exceed 36 N and eliminated results where this limit was reached. The FDA used a force machine with a limit of 1000 N, allowing for the testing of thicker home-based diets. The FDA also investigated two different rates of push: one fast (push rate of 12 ml/s) and one slow (1 ml/s). The quick push served as a worst case testing scenario and may not necessarily reflect common clinical practice (a slower push rate is more common). The Mayo Clinic used a single push rate of 1 ml/s. A final difference between the Mayo Clinic and the FDA pertains to how the data were combined. The FDA presented the data based upon the specific transition scenarios expected in the US, whereas the Mayo Clinic combined data from all legacy G-tubes and compared them against all ENFit-based G-tube devices. Trends observed by the two labs were similar. Apart from the same 14 Fr G-tube device that was found to be slower in gravity feed testing, the Mayo Clinic found that ENFit-based G-tube devices required a similar or a lower force relative to legacy devices, in most cases. The FDA, using the home-based diets, found that ENFit-based G-tube devices typically take less force (an example shown in Figure 5C) in 25 out of 27 scenarios, unless the diets were extremely thick (“thick” being qualitatively defined as visually having a peanut butter texture; viscosity values provided in Table S1), in which case ENFit-based G-tube devices required more force in a majority of the cases (18 out of 25 scenarios). The low-profile ENFit-based G-tube devices did not require more force compared with legacy devices. In this regard, low-profile G-tube devices are enteral devices with the distal section inside the stomach, similar to the standard G-tube devices; however, the visible proximal section is flush with the patient’s skin, which has the advantages of being able to be hidden under clothing. Low-profile G-tube devices are also considered to have a lower likelihood of dislodgement compared with standard G-tubes. The Japanese Pharmaceuticals and Medical Device Agency also conducted a similar force study,¹⁰ although their testing was performed on ENFit prototypes rather than finished devices. They used seven different commercial semisolid formulas and studied push rates of 0.83, 1.66, and 2.5 ml/s. Although the force associated with the prototype ENFit-based G-tube device was greater than legacy devices for a few diets (three out of seven) and device sizes (20 Fr), the absolute difference in force in most of these cases was <4–8 N, or 1–2 lbf, and the total force required was ~30–40 N, or ~7–9 lbf. It is not clear whether such differences would be clinically perceptible.

The clogging frequency of standard and low-profile ENFit-based G-tube devices was investigated by the FDA and compared with legacy devices.^12,19,22 Clogging was rare for all brands of standard legacy and ENFit-based G-tube devices when tested with commercial liquid diets under gravity (2 out of 96 cases for both legacy and ENFit-based G-tube devices).¹⁹ However, all standard ENFit-based G-tube device brands and some legacy G-tube device brands with narrow connector sections were more prone to clogging than two other standard legacy G-tube brands with a wider connector diameter when tested with gravity feeds and home-based or commercial blenderized diets.¹² For the blenderized diets that clogged the devices, using high-powered and industrial-grade blenders (~2000 W, cost $300–$500, brand names provided in Table S1) or pushing the diet with a force ranging from 8.8 to 18.2 N (depending on the device considered) eliminated this issue.¹² Low-profile legacy and low-profile ENFit-based G-tube devices were found to be comparably prone to clogging under gravity (18 out of 27 clogs for legacy compared with 19 out of 27 clogs for ENFit-based G-tube devices). They were also found to clog more frequently compared with standard G-tube devices. For syringe push-based feeding, clogging was rare for legacy and ENFit-based standard G-tube devices.

THE SCIENCE BEHIND THE FINDINGS

The findings discussed above can be explained by invoking energy loss theory, which has been described previously.²² Briefly, reduction in flow rate is strongly associated with reduced tube diameter and weakly associated with increased tube length. A standard G-tube can be divided into two broad sections: the connector section (near the medication port) and the distal section located inside the stomach. The outer diameter of the distal section is standardized by the device size (eg, 14 Fr or 24 Fr). With the standard G-tube devices that have been studied so far, the flow rate slowing can either be attributed to the smaller internal tube diameter of the ENFit-based G-tube device relative to its legacy counterpart (as was observed for one standard 14 Fr G-tube device in Mayo Clinic and FDA studies, E1 at 14 Fr in Figure 5B) or the narrowing in the connector section (as observed with some standard legacy and ENFit-based G-tube devices L1, L2, E1, and E2 at 24 Fr in Figure 5B). Given that the ENFit design’s inner diameter is 2.9 mm and is approximately equal to the inner distal diameter for most 18 Fr G-tube devices, it is likely that patients who currently use standard G-tube devices ≤18 Fr will not be impacted by the lower flow rate.

Unlike standard G-tube devices, low-profile G-tube devices are designed differently. As shown in Figure 4B, low-profile G-tube devices have the extension set (bolus or pump) that locks with the low-profile G-tube device only during feeding administration. Note that the ENFit connector for low-profile G-tube devices is on the extension set, and the low-profile G-tube devices themselves did not undergo any design changes (Figure 4B). In addition, the distal tube diameter for the extension sets (which remains outside the skin) is a standard size and does not change with various low-profile G-tube sizes. Thus, with no transition section²³ after the ENFit connector, and a short connector length (<10 mm), design changes in the extension sets of the low-profile G-tubes are minimal. Accordingly, there is not any perceptible loss in gravity flow rate transitioning from the legacy low-profile G-tube device with wider legacy bolus connector diameter (~4 mm) to the ENFit low-profile G-tube device with its connector’s 2.9 mm diameter. Consistent with this observation, most flow rates reported for low-profile G-tubes have been found to be similar to their legacy counterparts.^12,24

For the push mode of feeding, the syringe plays an important role in dictating the force required. The losses that need to be overcome in the push mode of feeding have two components: pushing the feed through the syringe and the G-tube device and overcoming the friction of plunger. We also know that the internal dimensional differences at 24 Fr vary across brands. Despite these differences, for the same diet and device size, the variability of force across different G-tube device brands is minimal. For example, in Figure 5C, the 24 Fr data show that the mean force is between 15 and 30 N in most cases across all brands. This implies that the energy loss in overcoming the plunger friction plays a larger role than the dimensional differences between different brands of G-tube devices. Like the flow rate testing data, the standard G-tube device L1 at 14 Fr (Figure 5C) was likely an outlier, as blenderized particulates may have obstructed the narrower flow pathways for this device. Note that both the FDA and the Mayo Clinic used only one brand of syringe with 60 ml volume with a tapered tip. Additional studies with different brands of syringes may help to better understand the impact of plunger friction and the tip designs on these reported results.

IMPACT OF TRANSITIONING TO ENFIT ON THE US TUBE FEEDING POPULATION

To estimate the fraction of the population affected by the transition from legacy to ENFit-based G-tube devices, we performed an analysis for this article using the FDA’s previously published experimental results^12,22 and a patient practices survey⁹ conducted by the Mayo Clinic. As a first step, a threshold characterizing a clinically relevant difference was established. For gravity feeding, in keeping with literature,²² the threshold was assumed to be 7 min out of a 20-min feeding, that is, a decrease in flow rate of 35% (7/20). Thus, when the reduction in flow rate exceeded 35%, the patient was considered to be affected. For syringe feeding, when the increase in push force in transitioning from a legacy to an ENFit-based G-tube device exceeded 2 lbf (ie, 8.9 N), the patient was assumed to be affected.¹² Using the FDA gravity flow rate²² and syringe push force¹² data, the mean and SD of the test measurements were used to create a normal distribution for the difference of the flow rate or force and the likelihood that the difference between legacy and ENFit-based G-tube devices exceeded the threshold was computed for different commercial and home-based diets, blenderized and nonblenderized diets, and G-tube types, brands, and sizes. To help quantify the uncertainty in our estimates, we performed a parametric study in which the thresholds that may potentially affect patients were decreased an additional 15%, 25%, 35%, and 50% for the flow rate and increased an additional +1 lbf, +2 lbf and +3 lbf for the push force. The procedure for converting the computed likelihoods to a level of impact for the tube feeding population, including the assumptions made and their limitations, is described further in the Supporting Information. The procedure for assessing the effect of uncertainty in marketing distribution, which involved an additional parametric study, is also described in the Supporting Information. The results of the calculations can be summarized as follows.

For adults using gravity feeding, we estimate that 2.5%–8.6% of the adult population may be affected by the transition to ENFit-based G-tube devices. For the gravity feeding pediatric population, 0.2%–1.9% of the population will be affected by this transition. For syringe push-based feeding, 1.5%–8.6% of the adult population and <0.5%–5.5% of the pediatric population will be affected by the transition to ENFit-based G-tube devices.

Although these results pertain only to standard and low-profile G-tube devices considered in the FDA and Mayo Clinic studies,^12,21 a wide variety of devices (14–24 Fr), device brands, and diets (including commercial, commercial blenderized, as well as home-based blenderized) were considered in the tests. Extrapolating the total number of enteral tube feeders to 531,832 in 2020, and assuming the proportion of adult and pediatric users remains the same as in 2013,⁸ we estimate that there are about 230,000 pediatric and 303,000 adult enteral users in the US. Translating the above percentages into total patients, we expect that about 600–4500 pediatric and 7700–26,200 adult patients using the gravity feeds and about 1100–12,700 pediatric and 4500–26,100 adult patients using the syringe push-based feeding could be impacted when transitioning to ENFit-based G-tube devices.

Enteral pumps are used by 28.4% and 17.6% of pediatric and adult populations, respectively.⁹ The impact calculations analyzed and presented in this review did not include patients using enteral pumps. However, enteral pumps deliver power through externally supplied energy and do not require physical effort when the diet is being delivered. The pumps also operate at specified flow rates independent of the G-tube device design. Therefore, the transition from legacy to ENFit-based G-tube devices is not expected to impact patients using enteral pumps. This is also supported by a recent study that compared the delivery of expressed breast milk for neonates using legacy and ENFit pump systems at different infusion rates.²⁵

SUGGESTIONS FOR STAKEHOLDERS

Our analysis of the performance testing, along with a consideration of the flow dynamics and energy losses in these devices, enables us to make the following observations and suggestions for various stakeholders:

Gravity feeds

1. Patients with low-profile G-tube devices are not likely to observe any changes when they transition to ENFit-based G-tube devices.

2. Patients with standard G-tube devices are likely to encounter slowing when transitioning to ENFit-based G-tube devices when delivering commercial liquid diets or using devices >18 Fr diameter with long transition sections. These patients may consider discussing with their physician if a larger-sized diameter connector/feeding tube is an option for them.

3. Patients using gravity feeds to deliver blenderized diets may also encounter slowing when transitioning to ENFit-based G-tube devices. These patients may consider discussing with their physician if a larger-sized diameter connector/feeding tube is an option for them. Such patients may also see increased clogging rates when transitioning from legacy to ENFit-based G-tube devices. Clogging can be mitigated by using high-powered blenders or strainers or by using diluents before blending diets. Since high-powered blenders can be cost prohibitive, another option for these patients is to use syringe push-based feeding.

4. The following strategies for increasing the flow rate through enteral devices should be considered for incorporation into standards: (i) designing devices to reduce the length of the connector section as well as to increase the inner diameters and (ii) evaluating with therapeutic nutrition enteral diets rather than water.

Syringe push-based feeding

1. Patients with low-profile G-tube devices are likely to find it easier to push their diets through their ENFit-based G-tube devices as opposed to gravity feeding.

2. Most patients with standard G-tube devices are not likely to require a significantly larger push force when they transition to ENFit-based G-tube devices. Patients using very thick diets may want to consider alternative thinner diets. Those patients who push their diets very rapidly (12 ml/s) may consider pushing slower.

3. Clogging can be reduced using high-powered blenders or strainers or by using diluents before blending diets.

4. The reduction of syringe plunger friction in ENFit-based syringes may be considered as an item to be incorporated into voluntary consensus standards.

DIRECTIONS FOR FUTURE RESEARCH

The studies considered in this paper did not address dosing accuracy for low-dose tip syringes,^11,15 cleaning comparisons between legacy and ENFit-based G-tube devices,²⁶ or the impact of microbial growth in connectors.²⁷ These areas, along with clinical research for gravity feeds and syringe push-based feeds, constitute important directions for future study.