Diagnostic interventions seek to determine whether GER may be the cause of pulmonary manifestations or failure to thrive. There is no single best test for “GERD” in infants. Instead, the clinician must combine specific testing with clinical context of a given patient. Several diagnostic tests are available.

pH-probe/impedance probe: probes introduce an esophageal tube with multiple side ports each able to detect pH or electrical impedance or both. For pH probes, a falling pH at a certain detector site in the esophagus is a proxy for gastric contents, presumably with low pH. Similarly, a decline in impedance suggests fluid at the detector. With multiple channels, these devices reliably detect GER (and its character—acidic, weakly acidic, or alkaline), and give an indication of its severity: how far up the esophagus refluxate travels, how often refluxate appears, and how fast it is cleared [24]. pH/impedance is often referred to as the “gold standard” for diagnosis of GER. It cannot however provide sole evidence that GER is actually GERD, nor can it reveal problems with gastric emptying or altered anatomy (like malrotation or hiatal hernia).

Fluoroscopy: To reveal anatomical problems that may present as GER, esophagosocopy and upper gastrointestinal (UGI) series are helpful. While insensitive to GER (reported sensitivity is under 50 %) [25], the images are far more specific in revealing hiatal hernia, H-type tracheoesophageal fistula, microgastria, esophageal stricture, and malrotation. Some surgeons insist on UGI before fundoplasty in order to plan the operation, for example to allow for the need to perform a Ladd’s procedure concomittantly [26] or even instead of fundoplication [27]. Data suggests that at least 4 % of patients being considered for fundoplasty will be found to have a surgically important abnormality [28]. But this value alone does not give the value of information from the UGI. For example it is reasonable to place premium value on decreased uncertainty about malrotation; that is, to assert that avoiding one missed malrotation is worth more than the cost of 24 negative studies. Value judgments of this kind resist standard calculus.

Manometry: this method uses a series of pressure sensors in the esophagus to describe esophageal pressure waves. Critical for diagnosis of motility disorders like achalasia in older children and adults, manometry contributes little to the workup of GERD in infants.

Esophagoscopy: Regarded as a mandatory diagnostic step by some, esophagoscopy is critical for diagnosis of esophageal metaplasia, ulcers, eosiniphilic esophagitis, conditions rarely encountered in newborns. Like manometry, esophagoscopy is not routinely used in newborns for workup of GERD.

Bronchoscopy: Bronchoscopy may provide compelling evidence of reflux with aspiration (see Fig. 2) but is not a front line tool in initial workup. On the other hand, patients who appear to have severe “reflux” may require interrogation of the airway to identify other structural problems that can mimic GERD, e.g., laryngeal cleft, H-type tracheoesophageal fistula, or tracheomalacia.

Nuclear scintigraphy: Radionuclide scintigraphy is attractive because it is noninvasive, carries no X-ray exposure, and can purportedly detect GER and microaspiration while quantifying gastric emptying. When compared to pH monitoring, the sensitivity of radionuclide scintigraphy is reported to be between 75 and 100 %, and specificity between 81.2 and 100 % [29]. But the usefulness of these “milk scans” in practice is less clear. Milk scans confirm that reflux is common in preterm infants, but it turns out that “positive scintigraphy has no correlation with symptoms” [30]. In that study, the authors found no discrimination at all in detection of GER between symptomatic babies and asymptomatic babies. Others report that the scans do not offer any information that guides whether pyloroplasty is indicated along with fundoplasty (Pyloroplasty is very rarely indicated in any case, and its routine use with fundoplication is condemned) [31]. Most likely, these scans are of little value in infants in the NICU: virtually always “positive” (even in children who have intact and working fundoplasty), their specificity appears considerably lower than reported for other populations. At the same time, the sensitivity to detect aspiration also appears in daily practice to be far below reported levels. In other words, in infants, they merely confirm reflux that is nearly universal, while failing to detect the “disease” in GERD.

Biomarkers: In patients undergoing bronchoscopy, who are intubated, or have a tracheostomy, bronchoalveolar lavage allows examination of the fluid in the airways. While traditional diagnosis relied on the presence of “lipid-laden macrophages” as a proxy for lung soilage, this finding appears to lack specificity [32]. Others have used the presence of pepsin [22, 33], measured according to various assays, to indicate GERD. This method seems to have promise, but lack of guidelines regarding interpretation of results, lack of standard methods for measurement, limited availability of the tests and restriction to patients with some tube in the airway all prevent widespread use for now.

N-of-one trials: While this method has the least published evidence, it is arguably the method to which experienced neonatal clinicians resort the most. N-of-1 trials are single-patient trials with “multiple-period crossover experiments comparing two or more treatments within individual patients” [34]. In detecting GERD, the clinician will observe the patient while being fed normally (orally or by tube) into the stomach. Frequency of apneas, oxygen requirement, discomfort, and other signs of symptomatic reflux are noted. Then feeds are withheld, with nutrition supported either by nasojejunal enteral feedings or parenteral nutrition. If GER is really GERD, improvement in respiratory and other symptoms will manifest within 2 or 3 days (sometimes less). Return to gastric feeds will reproduce the symptoms, and a second period of non-gastric nutrition relieves them. Each “block” of these trials is about 1 week long. While this method does not rule out laryngeal cleft or other anomalies in orally fed children, these confounders can be controlled by restricting enteral nutrition to tube feeding. While evidence is scant regarding this method, and criticisms are plain (jejunal feeds are imperfect protection against reflux; trials take a long time; etc.), this pragmatic method may yield the highest individual validity.

Traditionally, the mainstay of control of GERD has been drugs which fall into two categories: motility agents and acid blockers.

From the discussion above on pathophysiology, the usefulness of acid blockade on the control of GERD manifestations should seem suspect. Nothing in the function of the LES depends on pH; acid is weak in infants already; and decreased pH creates an important “deflector screen” against enteric organisms migrating from the stomach back into the airway. It is plausible that reliable acid control could act to deactivate pepsin and other digestive enzymes, but this has not been shown. At the same time, while improvements in pediatric asthma have been shown in the past, more recent trials show no benefit, and even harm in asthmatic children [35]. Meanwhile, there is increasing skepticism about acid blockade in infants. Proton-pump inhibitors (PPIs) carry serious risks and little effect on GERD symptoms [36], and even H2 blockers were recently associated with serious, even fatal, problems in infants [37]. While these medications have their place in gastritis with bleeding, or GERD manifesting largely as oral resistance and discomfort, it is plain that they are over-prescribed [1] and undereffective.

Motility agents fare little better. Metoclopramide (Reglan), bethanechol, cisapride, and erythromycin have all been tried as frontline treatments for GERD. In theory, increasing “motility” (i.e., decreasing gastric emptying time, decreasing intestinal transit time) should decrease GER events associated with TLESRs by reducing the dwell time of feeds in the stomach.

These posited effects have not been demonstrated practice. Reglan has no measurable effect on GERD in infants but carries a high risk of dystonic reactions [38]. Cisapride was shown to be effective, but was removed from the US market after several patients developed cardiac dysrythmias attributed to it [39]. Bethanechol, a parasympathomimetic, was posited to aid GERD by increasing intestinal motility, but trials have failed to demonstrated effectiveness (for example [40]). Erythromycin, a motilin agonist, certainly increases gastric emptying, but has not been shown to be an effective treatment for GERD, perhaps because of its strong tendency toward tachyphylaxis [39, 41]. At one time, it was common to place almost every baby in the NICU on “R&R” (reglan and ranitidine), a practice now, unsurprisingly, out of favor.

The reflux equation described above illustrates why other therapies that seem reasonable have unclear power to attenuate GER. Chief among these is the practice of “thickening” feeds. In this practice, a thickening agent (rice cereal, guar gum, cornstarch) is added to milk or infant formula. This “thickening” is actually a change in viscosity, η. From the equation, it is apparent that in order to halve GER, one would need to double the viscosity. The viscosity of human breast milk is around 3 cP (pumped, untreated), but addition of a commercial thickener (Nestle ThickenUp^®, a preparation of cornstarch) demonstrated a large increase in measured viscosity: “In relation to untreated pumped human milk without thickener, we observed that the addition of 7 % of the thickener increased the viscosity up to ninefold” [42]. So far so good. But this improvement comes at a cost—delayed gastric emptying. There are other costs: infants have diminished alpha-amylase and thus diminished ability to digest cornstarch which can produce diarrhea [43] Increasing viscosity delays gastric emptying [44]. With contrary effects and varied effects of different thickeners plus the confounder of varied viscosity of human infant milk formulas, mixed results with use of thickeners is unsurprising. Pectin, another thickening agent, appears to delay gastric emptying [45] but a trial in neurologically injured children showed improvement in GERD [46]. In sum, no reliable evidence exists to support or decry use of thickeners as GER treatment in infants. A Cochrane review in 2002 concluded “There is no evidence from randomised controlled trials to support or refute the efficacy of feed thickeners in newborn infants with [GER]” [47]. In a more recent study, the results are, again, mixed: “A formula thickened with amylopectin did not reduce the number of apnea of prematurity or GER-related apneas. It reduced acid GER features but had no effect on non-acid GER indexes” [48].

Others have turned to “elemental” formulas in order to treat GERD. There is nothing in these formulas that can be expected to improve the function of the LES or to increase gastric emptying. Rather the opposite is true—all of these formulas share one counterproductive characteristic, relatively high osmolarity. While breast milk has an osmolarity similar to serum (~290 mOsm/L), elemental formulas are much higher (e.g., Alimentum™, 370 mOsm/L when prepared at 20 kcal/30 mL). The stomach responds to elevated osmolarity by holding the fluid longer and diluting, since the small bowel does not tolerate high osmolarity fluid well (producing cramping, flushing, and diarrhea recognized as dumping syndrome). Changing a baby to an elemental formula may certainly be the right intervention for many diseases (protein hypersensitivity, malabsorption, etc.). But pathophysiologically, if a change to an elemental formula does ameliorate some observed manifestation of “feeding intolerance,” it is unlikely that the cause was GERD.

Surgical control of GER aims to restore normal functioning of the LES. “Normal” function does not include the inability to vomit, the inability to burp, or total elimination of reflux. Instead, surgery aims to re-establish or reinforce the mechanical functions that the LES normally employs [8]. To review:

An additional surgical objective is removal of barriers to gastric emptying (e.g., malrotation, overdistended stomach, etc.). The underlying objective is to attain these results without creating unwanted problems like dysphagia.

All surgical fundoplasties achieve these mechanical objectives. It may surprise readers familiar only with the “Nissen” fundoplication, but surgical control of GERD does not begin and end with the Nissen. Instead, there are several approaches (Table 1).

While all these procedures have an eponym, rarely does the modern application of various procedures conform to the original description. This chapter confines discussion to the three most common fundoplasty types performed in North America:

Nissen: the modern Nissen is sometimes called a “floppy Nissen” to distinguish it from its original namesake, or even its modified form the “Nissen–Rossetti.” The modern Nissen creates as 360° wrap looser and shorter (e.g., 2 cm) than originally described, and involves division of the short-gastric vessels between spleen and the greater curvature. Unlike the variant called a “Collis-Nissen,” the modern Nissen places the wrap entirely above the GEJ (Fig. 3).