Otolaryngology–Head and Neck Surgery (2006) 134, 81-85

ORIGINAL RESEARCH

Diagnosis of Laryngopharyngeal Reflux Among

Globus Patients
Kyung Hoon Park, MD, Sun Myung Choi, MD, Soon UK Kwon, MD,
Sang Won Yoon, MD, and Seon UK Kim, MD, Seoul, Korea
OBJECTIVE: Laryngopharyngeal reflux (LPR) is suggested as
one of the most common etiologic factors of globus; currently,
however, only a few tools are available to carry out effective
diagnosis of LPR in clinics. The goal of this study was to assess the
validity of reflux finding score (RFS), reflux symptom index (RSI),
and symptom checklist-90-revision (SCL-90-R) as diagnostic
methods for LPR.
STUDY DESIGN AND SETTING: We selected 57 globus
patients who were divided into test and control groups according to
the results of 24-hour double-probe ambulatory pH monitoring.
RFS, RSI, and SCL-90-R were then compared between the
2 groups, and the logical products of the 3 tests were also com-
pared.
RESULTS: RFS and RSI showed low specificity and there was
no significant difference between the test and control groups.
However, SCL-90-R revealed significantly high scores in the con-
trol group. The logical product of combining positive RFS and
negative SCL-90-R was significantly different between the groups.
CONCLUSION: RFS and RSI are less valid as diagnostic tools
for LPR when used independently. Patients with LPR are found to
exhibit weaker psychological symptoms than non-LPR patients.
EBM rating: B-2b
© 2006 American Academy of Otolaryngology–Head and Neck
Surgery Foundation, Inc. All rights reserved.
G lobus is the state of feeling a continuous sensation of a
lump or tightness in the throat. It accounts for 4% to 10%
of the complaints of patients who visit otolaryngology clinics,
and is reported to be on the increase in populations recently.1
However, the etiology is still unclear, which makes it difficult
for diagnosis and treatment. Nowadays, of the various causes
of globus, gastroesophageal reflux is reported to account for
23% to 68%, signifying its importance as a cause that should
be diagnosed properly among globus patients.2,3
Diagnostic methods for laryngopharyngeal reflux (LPR)
include 24-hour double-probe ambulatory pH monitoring,
gastroesophagogram, and gastroesophageal endoscopy. But
these are not easy to use in clinics, so there is a need for a
tool that is not only less invasive but also highly sensitive
and specific.4,5 Recently, other tools based on physical find-
ings and symptoms such as reflux finding score (RFS) and
reflux symptom index (RSI) have been tried in some clinics
and their simple, economic, and noninvasive characteristics
were reported.6,7
In this study, we performed 24-hour double-probe pH
monitoring, RFS, RSI, and the symptom checklist-90-revi-
sion (SCL-90-R)8 to globus patients and evaluated the di-
agnostic value of these tests for LPR.
MATERIALS AND METHODS
Fifty-seven patients (13 males and 44 females) who visited
the Otolaryngology Department at the Eulji Hospital from
April 2004 to July 2004 were selected. They complained of
throat symptoms such as sensation of a lump, something
sticking in the throat, or frequent throat clearing. Symptoms
should have continued more than a month. Patients who had
organic causes such as rhinitis, sinusitis, pharyngitis, laryn-
gitis, or tumorous lesions on history taking and physical
examination were excluded. The age distribution was from
23 to 73 years with the mean age of 46 (SD ⫽ 11.41). The
duration of symptom was in the range of 1 to 120 months
with a mean of 15 months (SD ⫽ 21.81). A detailed expla-

From the Department of Otolaryngology (Drs Park, Choi, Kwon, and ogy, Eulji Medical Center, #280-1 Hagye 1-Dong, Nowon-Gu, Seoul,
Yoon) and Psychiatry (Dr Kim), School of Medicine, Eulji University. Korea.
Reprint requests: Soon Uk Kwon, MD, Department of Otolaryngol- E-mail address: ksu11@netian.net.

0194-5998/$32.00 © 2006 American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. All rights reserved.
doi:10.1016/j.otohns.2005.08.025
82 Otolaryngology–Head and Neck Surgery, Vol 134, No 1, January 2006

Table 1 Table 3
Reflux finding score Symptom checklist-90-revision

Condition Score 9 symptom dimensions 3 global indexes

Subglottic edema 0 ⫽ absent Somatization

2 ⫽ present Obsessive-compulsive
Ventricular obliteration 2 ⫽ partial Interpersonal sensitivity
4 ⫽ complete Depression Global severity index
Erythema/hyperemia 2 ⫽ arytenoid only Anxiety Positive symptom total
4 ⫽ diffuse Hostility Positive symptom distress
Vocal cord edema 1 ⫽ mild Phobic anxiety index
2 ⫽ moderate Paranoid ideation
3 ⫽ severe Psychoticism
4 ⫽ polypoid
Diffuse laryngeal edema 1 ⫽ mild
2 ⫽ moderate
3 ⫽ severe cisely. In a diary, patients reported all meals and drinks and
4 ⫽ obstructing
P-commissure hypertrophy 1 ⫽ mild body position. The pH data were recorded on a Digitrapper
2 ⫽ moderate (Synetics Medical, Stockholm, Sweden) and analyzed using
3 ⫽ severe the EsopHogram software (Synetics Medical, Copenhagen,
4 ⫽ obstructing Denmark). All meal intervals were eliminated from analysis. A
Granuloma/granulation tissue 0 ⫽ absent single pharyngeal event (pH ⬍ 4.0) immediately preceded by
2 ⫽ present
Thick endolaryngeal mucus 0 ⫽ absent a precipitous drop in the esophageal probe of equal pH
2 ⫽ present magnitude is regarded as positive evidence for LPR. Ac-
cording to the results, patients were divided into test and
control groups.
Prior to this, at their first visits we had performed RFS,
nation of the study and procedures, including risks, benefits, RSI, and SCL-90-R. For RFS, an examiner uses a fiberoptic
and alternatives, was provided, and written informed con- laryngoscope to score 8-item scale describing physical find-
sent was obtained. However, an Institutional Review Board ings observed frequently in LPR patients. All RFS were
was not available. checked by the third author. A total score over 7 was
Before undergoing 24-hour double-probe ambulatory pH regarded to be LPR positive (Table 1). On the other hand,
monitoring, all patients underwent esophageal manometry for RSI, patients themselves give points to 9 symptoms that
using a station pull-through technique (Biomedicals, Ven- are frequently observed in LPR patients. A total score over
tura, CA) to identify the location of the lower and upper 13 was regarded as LPR positive (Table 2). SCL-90-R is a
esophageal sphincters relative to the nasal sill. Zinetics 24 psychological questionnaire composed of 90 items. As a
single-use pH catheters (Meditronic, Skovlunde, Denmark) result, we calculated the standardized T-scores for each of
with various intraprobe lengths from 10 cm to 21 cm were the 9 symptom dimensions and 3 global indexes (Table 3).
calibrated before monitoring. The distal pH probe was We analyzed the differences between the 2 groups with
placed in the distal esophagus 4 cm above the lower esoph- respect to their RFS, RSI, and SCL-90-R. Sensitivity and
ageal sphincter. The proximal pH probe was placed in the specificity of the diagnostic criteria determined by each test
hypopharynx 1 cm above the upper esophageal sphincter pre- were also evaluated. The Fisher’s exact test was used to

Table 2
Reflux symptom index

Within the last month, how did the following problems affect you? 0 ⫽ No problem
Circle the appropriate response. 5 ⫽ Severe problem

1. Hoarseness or a problem with your voice 0 1 2 3 4 5

2. Clearing your throat 0 1 2 3 4 5
3. Excess throat mucus or postnasal drip 0 1 2 3 4 5
4. Difficulty swallowing food, liquids, or pills 0 1 2 3 4 5
5. Coughing after you ate or after lying down 0 1 2 3 4 5
6. Breathing difficulties or choking episodes 0 1 2 3 4 5
7. Troublesome or annoying cough 0 1 2 3 4 5
8. Sensations of something sticking or a lump in your throat 0 1 2 3 4 5
9. Heartburn, chest pain, indigestion, or stomach acid coming up 0 1 2 3 4 5
Park et al Diagnosis of Laryngopharyngeal Reflux Among . . . 83

analyze the differences in RFS and RSI between the 2 groups.

The Student t test was used to analyze the differences in
SCL-90-R. Then the logical products of combining RFS,
RSI, or SCL-90-R were also analyzed as new diagnostic
criteria for LPR. These were evaluated using the Fisher’s
exact test. SPSS version 12.0 (SPSS, Inc., Chicago, IL) was
used for the analysis of data, where statistical significance
was determined if P ⬍ 0.05.

RESULTS
Differences in RFS Between the Two Groups
The test group identified as LPR positive by 24-hour dou-
ble-probe ambulatory pH monitoring included 41 patients
(71.9%). The control group identified as LPR negative had
16 patients (28.1%). On the other hand, those tested positive
for RFS were 46 (80.7%) and negative were 11 (19.3%).
Sensitivity and specificity of RFS were 87.8% and 37.5%,
respectively. Thus, RFS was relatively sensitive but showed
high false-positive rates (Table 4). Moreover, even if there
Table 4
Differences of RFS, RSI, SCL-90-R, and the logical
products of combining positive RFS, positive RSI,
or negative SCL-90-R between LPR patients and
non-LPR patients
LPR
Group ⫹ ⫺ Total P difference between the groups (P ⫽ 0.740).
Figure 1 Profile of symptom checklist-90-revision. *Significant
difference (P ⬍ 0.05). LPR, laryngopharyngeal reflux proved by
24-hour double-probe ambulatory pH monitoring; SOM, somatiza-
tion; O-C, obsessive-compulsive; I-S, interpersonal sensitivity;
DEP, depression; ANX, anxiety; HOS, hostility; PHOB, phobic
anxiety; PAR, paronoid ideation; PSY, psychoticism; GSI, global
severity index; PST, positive symptom total; PSDI, positive symp-
tom distress index.
were some differences in RFS between the groups, there
was no statistical significance (P ⫽ 0.057).
Differences in RSI Between the Groups
Forty-four patients (77.2%) tested positive and 13 patients
(22.8%) negative for RSI. Sensitivity and specificity of RSI
were 75.6% and 18.8%, respectively. Thus, RSI revealed
lower validity than RFS (Table 4). There was no significant

RFS Differences in SCL-90-R Between the Groups

⫹ 36 10 46
0.057 For the test group, the mean scores for each of the 9
⫺ 5 6 11
RSI symptom dimensions that included somatization (SOM),
⫹ 31 13 44 obsessive-compulsive (O-C), interpersonal sensitivity (I-S),
0.740
⫺ 10 3 13 depression (DEP), anxiety (ANX), hostility (HOS), phobic
SCL-90-R (PST)
anxiety (PHOB), paranoid ideation (PAR), and psychoti-
⫹ 37 11 48
0.099 cism (PSY) were 53.10, 45.24, 44.76, 46.71, 46.76, 46.17,
⫺ 4 5 9
SCL-90-R (PSDI) 46.56, 43.66, and 46.44, respectively. For the control, the
⫹ 35 14 49 mean values for the symptom dimensions in the same order
1.000
⫺ 6 2 8 were 55.13, 52.56, 52.56, 52.69, 53.63, 51.13, 53.94, 51.38,
RFS (⫹) & RSI (⫹)
and 53.13, respectively. The mean scores of global severity
⫹ 28 8 36
0.232 index (GSI), positive symptom total (PST), and positive
⫺ 13 8 21
RSI (⫹) & SCL-90-R (⫺) symptom distress index (PSDI) were 32.63, 46.76, and
⫹ 27 8 35 47.63 for the test group and 32.25, 53.38, and 55.12 for the
0.366
⫺ 14 8 22 control group (Fig 1). Both groups showed high SOM but
RFS (⫹) & SCL-90-R (⫺)
the control group revealed significantly higher scores than
⫹ 33 6 39
⫺ 8 10 18
0.003* the test group in all the other symptom dimensions (P ⬍
Total 41 16 57 0.05). Moreover, in the 3 global indexes, PST and PSDI
*Significant difference. SCL-90-R (⫺) was defined as a score
were significantly high in the control group (P ⬍ 0.05).
of PST below 60. RFS, reflux finding score; RSI, reflux symp- Defining scores of PST over 60 as positive SCL-90-R,
tom index; SCL-90-R, symptom checklist-90-revision; PST, we determined the negative SCL-90-R as a diagnostic cri-
positive symptom total; PSDI, positive symptom distress in- terion for LPR. Sensitivity and specificity of SCL-90-R
dex; LPR, laryngopharyngeal reflux proved by 24-hour dou- were 90.2% and 31.3%, respectively. On the other hand,
ble-probe ambulatory pH monitoring.
when we defined positive SCL-90-R as scores of PSDI over
84 Otolaryngology–Head and Neck Surgery, Vol 134, No 1, January 2006
60, sensitivity and specificity of negative SCL-90-R were
85.4% and 12.5%, respectively (Table 4). However, there
was no significant difference between the groups when
either PST or PSDI were used as diagnostic criteria (P ⫽
0.099, P ⫽ 1.000).
Differences in the Logical Product of
Combining Positive RFS and Positive RSI
Specificity was increased but sensitivity was decreased;
sensitivity and specificity were 68.3% and 50.0%, respec-
tively (Table 4). Moreover, there was no significant differ-
ence between the groups (P ⫽ 0.232).
Differences in the Logical Product of
Combining Positive RSI and Negative
SCL-90-R
First, the positive SCL-90-R was defined as a score of PST
over 60. The logical product of combining positive RSI and
negative SCL-90-R was used as a diagnostic criterion. As a
result, sensitivity and specificity were 65.9% and 50.0%,
respectively (Table 4), and significant difference was not
noted (P ⫽ 0.366).
Differences in the Logical Product of
Combining Positive RFS and Negative
SCL-90-R
Sensitivity and specificity were 80.5% and 62.5%, respec-
tively (Table. 4). It was the most valid result and there were
statistically significant differences between the groups (P ⫽
0.003).
DISCUSSION
In the past, globus was considered a kind of conversion or
psychosomatic disorder, but in recent decades, various or-
ganic or functional causes have been reported. Some of the
reported causes of globus are elongated styloid process,
rhinosinusitis, temporomandibular joint disorder, lingual
tonsillar hypertrophy, gastroesophageal reflux, and tumor-
ous lesion;9 among these, gastroesophageal reflux accounts
for 23% to 68% of all causes.2,3 In this study, 71.9% of all
patients who had no other organic causes had LPR.
Diagnostic methods for LPR include 24-hour double-probe
ambulatory pH monitoring, gastroesophagogram, and gastro-
esophageal endoscopy. The 24-hour double-probe ambulatory
pH monitoring has the most superior sensitivity and specificity,
so it is considered the confirmation test.4,5 However, it is not
usually used in clinics because it is relatively invasive and
restricts the ordinary activity of patients for a day. Moreover,
results of this test may be produced differently according to
examiners’ interpretation and false-positive or -negative results
may also be present.10,11
Gastroesophagogram and gastroesophageal endoscopy
may also contribute to the diagnosis of LPR. However,
these are seldom used in clinics because the laryngeal le-
sions that do not involve the esophagus cannot be detected
by these methods.5 Therefore, other tools that are not inva-
sive and have high validity are in need.
Recently, RFS and RSI have been used in some clin-
ics.6,7 RFS is an 8-item clinical severity scale that scores
physical findings, measuring in the range of 0 (no abnormal
findings) to a maximum of 26 (worst score possible). The
8 items of RFS are based on the most common laryngeal
findings in LPR patients. RSI is a self-administered 9-item
index based on the frequent symptoms in LPR patients. The
scale for each individual item ranges from 0 (no problem) to
5 (severe problem), with a maximum total score of 45. LPR
is diagnosed positive when RFS value is over 7 or when RSI
value is over 13; it is reported that RFS and RSI have
excellent validity and reproducibility. Moreover, these tests
take only about 1 minute each and are simple, economical,
and noninvasive. However, in the previous literature,
asymptomatic cases were defined as the control group and
their RFS and RSI were compared with those of the test
group having LPR. Therefore, it is doubtful that the use of
RFS and RSI can distinguish LPR patients from the rest of
the symptomatic patients who actually visit clinics. In our
research, we assigned those identified as nonreflux patients
by 24-hour double-probe ambulatory pH monitoring as the
control group and then studied the validity of RFS and RSI.
The result revealed some limitations of RFS and RSI as
diagnostic tools.
Although many reports have asserted organic causes for
globus patients, the possibility of psychological causes has
also been steadily investigated. Research investigating the
relationship between LPR and psychological factors, how-
ever, is rare. We analyzed the differences of the psycholog-
ical characteristics between LPR patients and non-LPR pa-
tients using SCL-90-R as a psychological tool.8 SCL-90-R
is a multidimensional self-report symptom inventory of 90
questions, each of which addresses one psychological
symptom. The result of this inventory comprises 9 symptom
dimensions (SOM, O-C, I-S, DEP, ANX, HOS, PHOB,
PAR, PSY) and 3 global indexes (GSI, PST, PSDI). Nine
symptom dimensions contain a comprehensive battery of
symptoms and show high correlation with the questions in
MMPI.12 Among the 3 global indexes, GSI represents the
depth of disorder and PST the numbers of present symp-
toms. PSDI implies the pure intensity of disorder and gives
information about patients’ response style to the symptoms.
SCL-90-R takes just 15 to 20 minutes and can be carried out
by amateurs. Thus it can be used as a screening test more
easily than other psychological tests.
In this study, globus patients tended to have somatization
regardless of LPR, but nonreflux patients revealed signifi-
cantly high scores at all the other symptom dimensions.
Among the 3 global indexes, PST and PSDI of nonreflux
patients were significantly high. This implies that nonreflux
patients tend to have more varied psychopathological prob-
lems and more subjective complaints. This result suggests
Park et al Diagnosis of Laryngopharyngeal Reflux Among . . .
that it is helpful to consider the psychological features for
the diagnosis of LPR.
Taking the above results into consideration, we at-
tempted to establish new diagnostic criteria by combining
RFS, RSI, or SCL-90-R. We found that the exclusion of
SCL-90-R-positive patients from those tested positive to
RFS serves as a relatively valid criterion for the diagnosis of
LPR. However, its reliability is not satisfying yet, so actual
pH monitoring is still necessary for the exact diagnosis.
CONCLUSION
For the diagnosis of LPR in globus patients, RFS and RSI
can be used as noninvasive, simple, and economical tools,
although they are less valid when used independently. SCL-
90-R reveals that fewer psychological symptoms and
weaker complaints are observed in LPR patients than in
non-LPR globus patients. Therefore, when used together,
RFS and SCL-90-R can serve as relatively valid tools for
10. Vincent DA Jr, Garrett JD, Radionoff SL, et al. The proximal probe in
the diagnosis of LPR in clinics.
11. Woo P, Noordzij P, Rosa JA. Association of esophageal reflux and
REFERENCES
12. Derogatis LR, Rickels K, Rock AF. The SCL-90 and the MMPI: a step
1. Mololy PJ, Charter R. The globus symptom. Arch Otolaryngol Head
Neck Surg 1982;108:740 – 4.
85
2. Wilson JA, Pryde A, Piris J, et al. Pharyngoesophageal dysmotility in
globus sensation. Arch Otolaryngol Head Neck Surg 1989;115:1086 –90.
3. Koufman JA, Amin MR, Panetti M. Prevalence of reflux in 113
consecutive patients with laryngeal and voice disorders. Otolaryngol
Head Neck Surg 2000;123:385– 8.
4. Koufman JA, Aviv JE, Casiano RR, et al. Laryngopharyngeal reflux:
position statement of the committee on speech, voice, and swallowing
disorders of the American Academy of Otolaryngology–Head and
Neck Surgery. Otolaryngol Head Neck Surg 2002;127:32–5.
5. Koufman JA. The otolaryngologic manifestations of gastroesophageal
reflux disease. A clinical investigation of 225 patients using ambula-
tory 24-hour pH monitoring and an experimental investigation of the
role of acid and pepsin in the development of laryngeal injury. Laryn-
goscope 1991;101:1– 65.
6. Belafsky PC, Postma GN, Koufman JA. The validity and reliability of
the reflux finding score (RFS). Laryngoscope 2001;111:1313–7.
7. Belafsky PC, Postma GN, Koufman JA. Validity and reliability of the
reflux symptom index (RSI). J Voice 2002;16:274 –7.
8. Derogatis LR. SCL-90 (revised) manual I. Clinical psychometrics
research unit. Baltimore: Johns Hopkins University School of Medi-
cine; 1977.
9. Timon C, O’Dwyer T, Cagney D, et al. Globus pharyngeus: long-term
follow-up and prognostic factors. Ann Otol Rhinol Laryngol 1991;
100:351– 4.
esophageal pH monitoring: development of a normative database. J
Voice 2000;14:247–54.
globus symptom: comparison of laryngoscopy and 24-hour pH ma-
nometry. Otolaryngol Head Neck Surg 1996;115:502–7.
in the validation of a new self-report scale. Br J Psychiatry 1976;128:
280 –9.