Abstract
PURPOSE
To evaluate the accuracy of signs and symptoms for the diagnosis of acute rhinosinusitis (ARS).
METHODS
We searched Medline to identify studies of outpatients with clinically suspected ARS and sufficient data reported to calculate the sensitivity and specificity. Of 1,649 studies initially identified, 17 met our inclusion criteria. Acute rhinosinusitis was diagnosed by any valid reference standard, whereas acute bacterial rhinosinusitis (ABRS) was diagnosed by purulence on antral puncture or positive bacterial culture. We used bivariate meta-analysis to calculate summary estimates of test accuracy.
RESULTS
Among patients with clinically suspected ARS, the prevalence of imaging confirmed ARS is 51% and ABRS is 31%. Clinical findings that best rule in ARS are purulent secretions in the middle meatus (positive likelihood ratio [LR+] 3.2) and the overall clinical impression (LR+ 3.0). The findings that best rule out ARS are the overall clinical impression (negative likelihood ratio [LR−] 0.37), normal transillumination (LR− 0.55), the absence of preceding respiratory tract infection (LR− 0.48), any nasal discharge (LR− 0.49), and purulent nasal discharge (LR− 0.54). Based on limited data, the overall clinical impression (LR+ 3.8, LR− 0.34), cacosmia (fetid odor on the breath) (LR+ 4.3, LR− 0.86) and pain in the teeth (LR+ 2.0, LR− 0.77) are the best predictors of ABRS. While several clinical decision rules have been proposed, none have been prospectively validated.
CONCLUSIONS
Among patients with clinically suspected ARS, only about one-third have ABRS. The overall clinical impression, cacosmia, and pain in the teeth are the best predictors of ABRS. Clinical decision rules, including those incorporating C-reactive protein, and use of urine dipsticks are promising, but require prospective validation.
Key words: sinusitis, rhinosinusitis, acute sinusitis, acute rhinosinusitis
INTRODUCTION
Acute rhinosinusitis (ARS) is defined as inflammation of the paranasal sinuses, most often the maxillary sinuses, that is caused by viruses or bacteria and has a duration of less than 6 weeks.1 Acute rhinosinusitis is a common outpatient infection, responsible for over 3 million outpatient visits annually in the United States; the symptoms overlap considerably with that of other upper respiratory tract infections, making accurate diagnosis challenging.2 While 75% of patients with ARS receive an antibiotic, and it is the most common reason for outpatient prescription of antibiotics,2,3 only about one-third with sinus symptoms have a confirmed bacterial pathogen when sinus fluid is cultured.4,5
Helping physicians more accurately identify which patients with clinically suspected sinusitis actually have acute bacterial rhinosinusitis (ABRS) could reduce harm from inappropriate antibiotic use. A systematic review found a 5% absolute increase in the rate of cure with antibiotics for clinically diagnosed ARS, compared with an 11% increase in rate of cure with imaging-diagnosed ARS; more accurate clinical diagnosis could identify the patients most likely to benefit from antibiotics.6 Previous systematic reviews of the clinical diagnosis of ARS are all more than 15 years old and did not use modern analytic techniques such as bivariate meta-analysis.7–10 The goal of the current study is therefore to perform a comprehensive meta-analysis of the clinical diagnosis of ARS and ABRS.
METHODS
Search Strategy
PubMed and Embase were searched using terms for acute rhinosinusitis and diagnostic testing (see Supplemental Table 1, available at http://www.AnnFamMed.org/content/17/2/164/suppl/DC1/, for search strategy).11 A separate search was performed to identify studies that assessed inter-rater agreement of signs and symptoms of sinusitis. PubMed was searched using the search terms (“inter-rater” OR “interrater” OR “kappa”) AND (“sinusitis” OR “sinus”). The primary searches occurred in 2017, and were updated in April, 2018.
Inclusion Criteria And Quality Assessment
We included studies of adults and children with suspected ARS or acute respiratory tract infection that reported data for the accuracy of at least 1 sign or symptom. One study12 included patients with a clinical diagnosis of ARS for whom their physician recommended antibiotics, but the results were similar to those of studies with broader inclusion criteria. All studies took place in the outpatient setting and used radiography, ultrasound, computed tomography (CT), or antral puncture as the reference standard. We included studies in which all patients received the same reference standard. Studies involving highly specialized patient populations (ie, patients with HIV or odontogenic sinusitis, children with brain cancer, or inpatients) were excluded. There were no limits by date of publication or language.
Where studies reported findings separately by maxillary, frontal, or ethmoid sinus, only maxillary sinus findings are shown. Where individual sinuses as well as results by person are reported, results by person are shown. When it was possible to use different cut points (definitions of abnormal) for a test or reference standard, the cut point that yielded the highest diagnostic odds ratio (DOR) was selected.13,14 Where data were reported for 2 time periods (ie, any fever or fever in the last 24 hours) we included the most recent time period in relation to the visit. Where results for more than 1 reference standard were reported for the same set of patients, data for the highest quality reference standard are reported (in descending order: positive bacterial culture of antral puncture fluid, antral puncture revealing purulent fluid, magnetic resonance imaging [MRI], CT, ultrasound, and finally radiography).4,15–18 Acute rhinosinusitis was diagnosed when any reference standard was abnormal, and acute bacterial rhinosinusitis when inspection of antral puncture fluid or culture of puncture fluid were consistent with bacterial infection.4,5
To evaluate study quality, we used the Quality Assessment of Diagnostic Accuracy Studies 2 criteria, adapted for the diagnosis of acute rhinosinusitis (Supplemental Table 2).19 Studies were classified as low risk of bias if all 4 domains were all judged to be low risk of bias. Studies with only a single domain at high risk of bias were classified as moderate risk of bias, and all other studies were classified as high risk of bias.
Data Extraction and Analysis
Each included study was reviewed by 2 investigators, who extracted data regarding study quality and the accuracy of signs and symptoms. Any discrepancies were resolved by discussion to achieve consensus, involving a third investigator if necessary. We used the MADA package in R version 3.2.2 (R Project for Statistical Computing) to perform bivariate meta-analysis for each clinical sign or symptom, and the META package to calculate summary estimates of prevalence. We determined the prevalence of sinusitis for subgroups by age and reference standard using a random effects summary measure. Summary measures of accuracy are reported for each sign or symptom. The positive likelihood hood ratios (LR+) and negative likelihood ratios (LR−) were the primary measures of diagnostic accuracy. A likelihood ratio (LR) near 1.0 means that the test adds little diagnostic information, LRs greater than 1 increase the likelihood of disease, and LRs less than 1 decrease the likelihood of disease.20 The DOR (LR+ divided by LR−) was chosen as an overall measure of discrimination because the small numbers of studies made receiver operating characteristic curves unstable and difficult to interpret in many cases.
RESULTS
We identified 1,638 studies after removing duplicates. We also searched the reference lists of previous meta-analyses, review articles, and practice guidelines for additional articles, finding an additional 11 studies. A total of 1,649 abstracts were screened by 2 reviewers for relevance, 182 full-text articles were accessed, and a final total of 17 studies met our inclusion criteria (Supplemental Figure 1).
Study Characteristics
The characteristics of the 17 included studies are summarized in Table 1. Six were small (fewer than 100 participants), with a range of 30 to 400 participants. The Quality Assessment of Diagnostic Accuracy Studies 2 framework is summarized in Supplemental Table 2. Overall, 4 studies were at low risk of bias, 7 at moderate risk of bias, and 6 at high risk of bias. Common threats to validity included failure to use a high-quality reference standard, nonconsecutive sample of patients, and mask the person performing the reference standard to results of the index test. All of the studies at low risk of bias used purulent fluid or bacterial culture as the reference standard, so those results are reported separately.33
Table 1.
Characteristics of Included Studies (N = 17)
| Study | Population | Age Data | Reference Standard | Country | Risk of Bias |
|---|---|---|---|---|---|
| Referral setting (ENT, respiratory, or allergy clinic) | |||||
| McNeill,13 1963 | Adults and children (n = 150, 242 sinuses) referred to ENT clinic for clinically suspected ARS | Range ≥10 y (10-19 y, n = 22; ≥20 y, n = 128) | Radiography showing mucosal thickening or any opacity | Northern Ireland | High |
| Axelsson et al,21 1976 | Consecutive adults (n = 164) at ENT clinic with clinically suspected acute maxillary sinusitis | Mean 35 y | Radiography (at least 4 views) showing any mucosal thickening, air-fluid levels, or opacification | Sweden | Moderate |
| Berg et al,15 1981 | Adults (n = 50) at ENT clinic with clinically suspected ARS of ≥3 weeks duration | Mean 46 y | Antral puncture revealing purulent discharge | Sweden | High |
| Berg et al,22 1985 | Adults (n = 90) at ENT clinic with clinically suspected ARS of ≥3 weeks duration | Not reported | Antral puncture revealing purulent discharge | Sweden | Moderate |
| van Buchem et al,5 1995 | Adults (n = 113) referred to ENT clinic with clinically suspected acute maxillary sinusitis | 42% 18-29 y; 34% 30-44 y; 16% 45-59 y; and 9% ≥60 y | (1) Antral puncture showing fluid or floccules (by patient) and (2) bacterial culture of fluid (by sinus) | Netherlands | Low |
| Visca et al,23 1995 | Children (n = 30) at pediatric respiratory clinic with clinically suspected ARS | Range 5-15 y | CT scan abnormal in coronal projection | Italy | High |
| Huang et al,24 2008 | Consecutive adults and children (n = 217) at allergy clinic with clinically suspected ARS of <3 weeks duration | Range 4-61 y (4-9 y, n = 89; 10-19 y, n = 101; ≥20 y, n = 27) | Sinus radiograph (n = 151) or CT scan (n = 12) with >4 mm mucosal thickening, air-fluid levels, and/or increased opacity or retention cyst | USA | High |
| Primary care, urgent care, or emergency department setting | |||||
| Berg et al,25 1988 | Adults (n = 155) presenting to ED with clinically suspected maxillary sinusitis of <3 months duration | Mean 38 y | Antral puncture with return of purulent or cloudy fluid | Sweden | Moderate |
| Williams et al,26 1992 | Consecutive men (n = 247) presenting to VA general medicine clinic with <3 months of self-described sinusitis or at least 1 sinus symptom–median symptom duration of 11 days | Median 50 y, IQ range 40-63 y | Radiographs (4 views): mucosal thickening ≥6 mm, complete opacity, or air-fluid level | USA | Moderate |
| van Duijn et al,27 1992 | Adults (n = 400, 441 episodes) presenting to primary care clinic with clinically suspected ARS | Range ≥15 y | Ultrasonography abnormal | Netherlands | Moderate |
| Hansen et al,4 1995 | Consecutive adults (n = 174) at primary care clinic suspected of having acute maxillary sinusitis by their general practitioner | Median 35 y, range 18-65 y | CT scan abnormal and culture of purulent fluid positive for pathogenic bacteria | Denmark | Low |
| Lindbaek et al,12 1996 | Adults (n = 201) clinically diagnosed by primary care doctor with ARS requiring antibiotics | Mean 37.8, range 15-83 y | CT scan showing air-fluid level or complete opacification | Norway | High |
| Laine et al,28 1998 | Consecutive adults (n = 39) presenting to primary care clinic with clinically suspected acute maxillary sinusitis of <30 days duration | Median 37 y, range 16-68 y | Nasal aspirate with purulent or mucopurulent material | Finland | Low |
| Varonen et al,29 2003 | Consecutive adults (n = 148) presenting to a primary care clinic with clinically suspected ARS of <30 days duration, 72% >5 days | Mean 39.7 y, range 18-75 y | Sinus radiographs (AP and Waters’ views) showing total opacification, air-fluid level, or mucosal thickening ≥6 mm | Finland | High |
| Thomas et al,30 2006 | Adults (n = 60) presenting to a VA urgent care center with clinically suspected ARS of <4 weeks duration | Mean 51 y, range 25-83 y | CT scan showing air-fluid level or complete opacification (mucosal thickening alone was not considered diagnostic) | USA | Moderate |
| Shaikh et al,31 2013 | Children (n = 258) at a general pediatric clinic with clinically suspected ARS of <30 days durations (mean symptom duration 14 days) | Mean 6.4 y, range 2.0-12.9 y | Radiographs (AP and Water’s views): complete opacification or any mucosal thickening | USA | Moderate |
| Autio et al,32 2015 | Adults (n = 50) presenting to a military clinic with clinically suspected ARS of <4 days duration (symptoms recorded at 9-10 days after onset) | Mean 20 y, range 18-23 y | CT followed by antral puncture and bacterial culture if positive | Finland | Low |
AP = anteroposterior; ARS = acute rhinosinusitis; CT = computed tomography; ED = emergency department; ENT = ear, nose, and throat; IQ = interquartile; USA = United States of America; VA = Veterans Administration.
Prevalence of ARS and ABRS
The prevalence of acute sinusitis for patients presenting with sinus symptoms is summarized in Table 2, stratified by age group and reference standard. Prevalence ranged from 19% to 63% for adults, and from 57% to 79% for children.
Table 2.
Prevalence of ARS and ABRS in the Included Study Populations
| Study | Reference Standard | No. | Prevalence, % |
|---|---|---|---|
| Studies of adults | |||
| Berg et al,15 1981 | Antral puncturea | 50 | 50 |
| Berg et al,22 1985 | Antral puncturea | 90 | 48 |
| Berg et al,25 1988 | Antral puncturea | 155 | 44 |
| Hansen et al,4 1995 | Antral puncturea | 174 | 53 |
| Van Buchem et al,5 1995 | Antral puncturea | 108 | 49 |
| Laine et al,28 1998 | Antral puncturea | 39 | 51 |
| Summary for antral puncture, % (95% CI) | 49 (45-53) | ||
| Hansen et al,4 1995 | Bacterial cultureb | 174 | 35 |
| Van Buchem et al,5 1995 | Bacterial cultureb | 113 | 34 |
| Autio et al,32 2015 | Bacterial cultureb | 43 | 19 |
| Summary for bacterial culture, % (95% CI) | 31 (24-40) | ||
| Lindbaek et al,12 1996 | CT | 201 | 63 |
| Thomas et al,30 2006 | CT | 60 | 45 |
| Van Duijn et al,27 1992 | Ultrasound | 423 | 49 |
| Varonen et al,29 2003 | Radiograph | 32 | 41 |
| Williams et al,26 1992 | Radiograph | 247 | 39 |
| Axelsson et al,21 1976 | Radiograph | 164 | 63 |
| Summary for imaging, % (95% CI) | 51 (48-54) | ||
| Summary for studies with adults only, % (95% CI) | 48 (42-54) | ||
| Studies of adults and children | |||
| McNeill,13 1963 | Radiograph | 242 | 56 |
| Huang et al,24 2008 | Radiograph | 217 | 70 |
| Summary for studies with adults and children, % (95% CI) | 63 (49-75) | ||
| Studies of children | |||
| Visca et al,23 1995 | CT | 30 | 57 |
| Shaikh et al,31 2013 | Radiograph | 258 | 79 |
| Summary for studies with children only, % (95% CI) | 70 (45-87) |
ABRS = acute bacterial rhinosinusitis; ARS = acute rhinosinusitis; CT = computed tomography.
A positive antral puncture required return of purulent fluid.
Bacterial culture of fluid obtained during antral puncture.
Imaging studies had the highest prevalence and was similar for studies using plain film radiography or CT as the reference standard (59% vs 56%, P = .70). For the diagnosis of ABRS, studies using the presence of purulence from antral puncture had a higher prevalence than those using positive culture of antral puncture fluid (49% vs 31%, P <.01). Comparing the studies of adults only (Table 2), the prevalence of ARS was similar for studies using any imaging vs purulence on antral puncture as the reference standard (51% vs 49%), but was lower for those using bacterial culture (31%).
In the subset of all primary care, urgent care, or emergency department studies (n = 10; 1,632 patients) the prevalence of ARS was 49% (95% CI, 39-59). In the subset of studies using antral puncture or culture as the reference standard (n = 4; 411 patients), the prevalence of ABRS was 42% (95% CI, 31-55).
All studies enrolled patients with symptoms compatible with sinusitis. In a systematic review of blood tests and imaging for ARS,11 2 studies that recruited patients with broader inclusion criteria of cold18 or runny nose34 found lower ARS prevalence at 16%and 28%, respectively.
The Interrater Reliability of Signs and Symptoms
Three studies reported the interrater agreement of signs and symptoms for ARS among adult patients.26,30,35 A value of κ from 0.0-0.2 represents slight agreement, 0.2-0.4 is fair agreement, 0.4-0.6 is good agreement, 0.6-0.8 substantial agreement, and 0.8-1.0 shows near perfect agreement. There was substantial agreement for history of cough (κ = 0.70),35 red streak in lateral recess of oropharynx (κ = 0.70),30 colored nasal discharge (κ = 0.68),26 and maxillary toothache (κ = 0.60).26 There was good agreement for sinus tenderness (κ = 0.59)26 and history of fever (κ = 0.53).35 There was only slight agreement for purulence on nasal inspection (κ = 0.14),26 and considerable heterogeneity from 2 studies regarding sinus transillumination (κ values of 0.22 and 0.80).26,30
Accuracy of Individual Signs and Symptoms Acute Rhinosinuitis
The individual signs and symptoms that best ruled in ARS when present were purulent secretions observed in the middle meatus, nasal speech, patient report of pain in the teeth, and purulent secretions observed in the posterior pharynx or postnasal region. The individual findings that best ruled out ARS were absence of a preceding respiratory tract infection, absence of any nasal discharge, absence of purulent nasal discharge as a symptom, and normal transillumination. The accuracy of individual signs and symptoms for the diagnosis of ARS is summarized in Table 3 (see Supplemental Table 3, for complete individual study data). Few positive findings had a LR ≥2.0 and few findings when absent had a LR ≤0.5.
Table 3.
Accuracy of Signs and Symptoms of ARS Diagnosed by Any Reference Standard
| Sign or Symptom | # of Studies | Sensitivity (95% CI) | Specificity (95% CI) | LR+ (95% CI)a | LR− (95% CI)a | DOR (95% CI)b |
|---|---|---|---|---|---|---|
| Overall clinical impression | 4 | 0.73 (0.66-0.79) | 0.75 (0.64-0.84) | 3.0 (2.1-4.4)c | 0.37 (0.29-0.46)c | 8.3 (4.9-13.1)c |
| Symptoms | ||||||
| Nasal speech | 3 | 0.51 (0.15-0.85) | 0.73 (0.35-0.93) | 1.9 (1.3-2.7)c | 0.67 (0.37-0.91)c | 2.9 (1.8-4.3)c |
| Pain in teeth | 8 | 0.34 (0.21-0.50) | 0.80 (0.63-0.90) | 1.8 (1.2-2.5)c | 0.83 (0.74-0.90)c | 2.1 (1.4-3.0)c |
| Nasal discharge | ||||||
| Purulent | 10 | 0.71 (0.63-0.78) | 0.54 (0.48-0.61) | 1.6 (1.4-1.7)c | 0.54 (0.44-0.56)c | 2.9 (2.2-3.7)c |
| Any | 4 | 0.75 (0.57-0.91) | 0.49 (0.35-0.68) | 1.5 (1.3-1.8)c | 0.49 (0.24-0.82)c | 3.5 (1.7-6.2)c |
| Maxillary or frontal pain | 3 | 0.49 (0.15-0.83) | 0.63 (0.14-0.95) | 1.50 (0.86-3.50) | 0.89 (0.61-1.40) | 1.90 (0.61-4.50) |
| Maxillary pain | ||||||
| Unilateral | 5 | 0.30 (0.14-0.52) | 0.80 (0.51-0.94) | 1.50 (0.96-2.60) | 0.90 (0.79-1.00)c | 1.70 (0.92-3.00) |
| Location not specified | 5 | 0.70 (0.42-0.88) | 0.20 (0.05-0.51) | 0.88 (0.73-1.00)c | 1.60 (0.95-2.80) | 0.60 (0.30-1.10) |
| Cacosmiad | 5 | 0.23 (0.15-0.33) | 0.84 (0.58-0.95) | 1.50 (0.63-3.70) | 0.96 (0.83-1.20) | 1.70 (0.53-4.30) |
| Double sickening | 3 | 0.74 (0.32-0.95) | 0.41 (0.13-0.77) | 1.30 (0.77-2.40) | 0.69 (0.20-1.50) | 2.70 (0.49-8.50) |
| Hyposmia or anosmia | 8 | 0.61 (0.52-0.70) | 0.54 (0.46-0.61) | 1.3 (1.1-1.5)c | 0.72 (0.59-0.87)c | 1.9 (1.3-2.5)c |
| Cough | ||||||
| Nocturnal | 3 | 0.89 (0.80-0.94) | 0.11 (0.06-0.19) | 0.99 (0.87-1.10) | 1.20 (0.43-2.60) | 1.10 (0.35-2.60) |
| Any | 7 | 0.69 (0.57-0.80) | 0.40 (0.27-0.55) | 1.3 (1.0-1.7)c | 0.78 (0.60-0.98)c | 1.5 (1.0-2.2)c |
| Preceding respiratory tract infection | 5 | 0.87 (0.66-0.96) | 0.27 (0.13-0.48) | 1.2 (1.1 -1.3)c | 0.48 (0.30-0.72)c | 2.6 (1.6-4.0)c |
| Pain bending forward | 5 | 0.72 (0.56-0.84) | 0.39 (0.24-0.57) | 1.2 (1.0-1.5)c | 0.71 (0.50-0.97)c | 1.70 (0.97-2.70) |
| Nasal congestion or obstruction | 9 | 0.83 (0.74-0.89) | 0.24 (0.14-0.36) | 1.1 (1.0-1.2)c | 0.73 (0.56-0.93)c | 1.5 (1.1-2.1)c |
| Fatigue or malaise | 6 | 0.62 (0.51-0.71) | 0.45 (0.32-0.59) | 1.10 (0.99-1.30) | 0.86 (0.72-1.00)c | 1.30 (0.98-1.80) |
| Sneezing | 3 | 0.63 (0.49-0.74) | 0.39 (0.31-0.48) | 1.00 (0.86-1.20) | 0.96 (0.71-1.20) | 1.10 (0.70-1.20) |
| Headache | 6 | 0.50 (0.32-0.67) | 0.50 (0.27-0.73) | 1.00 (0.71-1.30) | 1.00 (0.79-1.40) | 1.00 (0.57-1.70) |
| Allergies by history | 3 | 0.19 (0.16-0.24) | 0.79 (0.72-0.85) | 0.96 (0.64-1.40) | 1.00 (0.93-1.10) | 0.96 (0.57-1.50) |
| Postnasal drip | 3 | 0.74 (0.55-0.87) | 0.19 (0.07-0.45) | 0.94 (0.73-1.30) | 1.50 (0.64-3.40) | 0.80 (0.23-2.00) |
| Previous sinusitis | 4 | 0.59 (0.52-0.65) | 0.27 (0.18-0.39) | 0.81 (0.71-0.93)c | 1.6 (1.1-2.2)c | 0.55 (0.33-0.83)c |
| Signs | ||||||
| Purulent secretions | ||||||
| Middle meatus | 5 | 0.15 (0.07-0.28) | 0.95 (0.87-0.98) | 3.2 (1.4-6.6)c | 0.90 (0.78-0.97)c | 3.7 (1.5-7.6)c |
| Pharyngeal or postnasal | 7 | 0.14 (0.06-0.30) | 0.92 (0.77-0.97) | 1.8 (1.1-2.8)c | 0.93 (0.87-0.98)c | 2.0 (1.2-3.1)c |
| Any nasal | 12 | 0.39 (0.28-0.51) | 0.74 (0.63-0.82) | 1.50 (0.99-2.30) | 0.84 (0.67-1.00)c | 1.90 (0.98-3.30) |
| Transillumination abnormal | 3 | 0.69 (0.63-0.75) | 0.56 (0.50-0.62) | 1.6 (1.4-1.9)c | 0.55 (0.44-0.67)c | 3.0 (2.0-4.2)c |
| Sinus tenderness | ||||||
| Frontal | 3 | 0.25 (0.09-0.54) | 0.75 (0.47-0.91) | 1.10 (0.76-1.40) | 0.99 (0.86-1.10) | 1.10 (0.70-1.60) |
| Maxillary | 12 | 0.44 (0.28-0.61) | 0.49 (0.34-0.64) | 0.88 (0.57-1.30) | 1.20 (0.80-1.60) | 0.81 (0.37-1.30) |
| Nasal mucosal thickening | 3 | 0.82 (0.27-0.98) | 0.30 (0.15-0.51) | 1.10 (0.49-1.40) | 0.68 (0.09-1.60) | 3.00 (0.32-13.00) |
| Temperature >38° C | 7 | 0.11 (0.07-0.19) | 0.87 (0.79-0.92) | 0.94 (0.42-1.90) | 1.00 (0.91-1.10) | 0.94 (0.37-2.10) |
ARS = acute rhinosinusitis; DOR = diagnostic odds ratio; LR+ = positive likelihood ratio; LR− = negative likelihood ratio.
Note: Each sign or symptom was reported in 3 or more studies to be included here.
A likelihood ratio near 1.0 means that the test adds little diagnostic information, a likelihood ratio >1 increases the likelihood of disease, and a likelihood ratio <1 decreases the likelihood of disease.
The diagnostic odds ratio is LR+ divided by LR− and is an overall measure of diagnostic discrimination.
Value is significantly < or > 1.0.
Cacosmia is fetid odor on patient’s breath.
Four studies reported data for the overall clinical impression as a diagnostic test for ARS, including 3 that used antral puncture as the reference standard.15,22,25,26 The accuracy of the overall clinical impression in this study was good (LR+ 3.0, LR− 0.37), with the highest diagnostic odds ratio of any finding (DOR 8.3).
Three studies reported data for prolonged duration of symptoms and the likelihood of ARS, using cutoffs of 5, 10, and 21 days.25,29,30 In our study, there was no clear pattern, with sensitivity ranging from 25% to 70% and specificity from 15% to 75%.
Acute Bacterial Rhinosinusitis
Six studies used purulent antral puncture fluid or positive bacterial culture as the reference standard and are summarized in Table 4.4,5,15,25,28,32 The 3 findings that significantly increased or decreased the likelihood of ABRS were the overall clinical impression, cacosmia (fetid odor of the breath), and pain in the teeth. Individual study data using these reference standards is shown in Supplemental Table 4. No conclusions can be drawn due to the limited number of studies reporting data for each sign or symptoms.
Table 4.
Accuracy of Signs and Symptoms of ABRS Diagnosed by Purulent Antral Puncture or Positive Bacterial Culture
| Sign or Symptom | # of Studies | Sensitivity (95% CI) | Sensitivity (95% CI) | LR+ (95% CI)a | LR− (95% CI)a | DOR (95% CI)b |
|---|---|---|---|---|---|---|
| Overall clinical impression | 3 | 0.74 (0.61-0.84) | 0.80 (0.72-0.87) | 3.9 (2.4-5.9)c | 0.33 (0.20-0.50)c | 13.0 (5.0-27)c |
| Symptoms | ||||||
| Cacosmiad | 3 | 0.23 (0.11-0.42) | 0.93 (0.59-0.99) | 4.3 (0.94-14) | 0.86 (0.76-0.99)c | 5.0 (0.74-18) |
| Pain in teeth | 3 | 0.38 (0.10-0.78) | 0.80 (0.37-0.97) | 2.0 (1.1-3.7)c | 0.77 (0.50-0.96)c | 2.7 (1.3-4.6)c |
| Purulent nasal discharge | 3 | 0.64 (0.45-0.79) | 0.50 (0.36-0.63) | 1.3 (0.99-1.6) | 0.74 (0.47-1.0) | 1.8 (0.98-3.2) |
| Pain bending forward | 3 | 0.64 (0.4-0.79) | 0.38 (0.20-0.60) | 1.0 (0.87-1.3) | 0.95 (0.69-1.3) | 1.2 (0.69-1.8) |
| Nasal congestion or obstruction | 4 | 0.71 (0.57-0.82) | 0.25 (0.10-0.52) | 0.98 (0.83-1.3) | 1.3 (0.70-2.3) | 0.88 (0.38-1.8) |
| Cough | 3 | 0.66 (0.34-0.86) | 0.29 (0.12-0.55) | 0.93 (0.70-1.1) | 1.2 (0.79-1.7) | 0.84 (0.45-1.4) |
| Physical examination | ||||||
| Sinus tenderness, maxillary | 5 | 0.33 (0.13-0.61) | 0.57 (0.40-0.73) | 0.86 (0.23-2.2) | 1.2 (0.55-2.1) | 0.97 (0.11-3.9) |
| Nasal discharge, purulent | 5 | 0.27 (0.20-0.35) | 0.66 (0.45-0.82) | 0.88 (0.36-1.9) | 1.2 (0.57-1.8) | 0.83 (0.22-2.3) |
ABRS = acute rhinosinusitis; DOR = diagnostic odds ratio; LR+ = positive likelihood ratio; LR− = negative likelihood ratio.
Note: Each sign or symptom was reported in 3 or more studies to be included here.
A likelihood ratio near 1.0 means that the test adds little diagnostic information, a likelihood ratio >1 increases the likelihood of disease, and a likelihood ratio <1 decreases the likelihood of disease.
The diagnostic odds ratio is LR+ divided by LR− and is an overall measure of diagnostic discrimination.
Value is significantly < or > 1.0.
Cacosmia is fetid odor on patient’s breath.
Only 2 studies reported the accuracy of signs and symptoms for ABRS using bacterial culture of antral fluid as the reference standard.4,32 Pain in teeth as a symptom (LR+ 2.8, LR− 0.76) and maxillary tenderness on exam (LR+ 1.8, LR− 0.71) significantly changed the likelihood of ABRS.
Accuracy of Combinations of Signs and Symptoms
Acute Rhinosinusitis
Four studies reported combinations of findings for the diagnosis of ARS (Table 5).12,24,26,27 Lindbaek and colleagues proposed a 4-item score; patients with all 4 findings present had a LR of 25 for ARS, those with 2 or 3 findings a LR of 1.2, and those with 0 or 1 finding had a LR of 0.16.12 Williams and colleagues developed a 5-point clinical score using maxillary toothache, abnormal transillumination, poor response to nasal decongestants or antihistamines, colored nasal discharge, or mucopurulence on examination for the diagnosis of ARS using radiography as the reference standard; likelihood ratios ranged from 6.4 for 4 or 5 points to 0.16 for 0 points.26
Table 5.
Combinations of Findings and Clinical Decision Rules for the Diagnosis of Acute Sinusitis
| Study | Reference Standard | Findings | Interpretation
|
||
|---|---|---|---|---|---|
| Result | Likelihood Ratio (95% CI) | Probability of Sinusitis, %a | |||
| Diagnosis of ARS | |||||
| Huang et al,24 2008b | Radiograph (n = 205) or CT (n = 12) | Leukocyte esterase: <1+ = 0, 1+ = 2, ≥2+ = 3 | 4-11 | 127.0 (8.1-2016.0) | 99 |
| 2-3 | 0.23 (0.10-0.52) | 13 | |||
| pH: <7.5 = 0, 7.5 = 1, 8.0 = 2, 8.5 = 3 Nitrite: none = 0, light pink = 1, dark pink = 2 Protein: <2+ = 0, 2+ = 1, 3+ = 2, 4+ = 3 |
0-1 | 0.00 (0.00-0.06) | 0 | ||
| Lindbaek et al,12 1996 | CT | Double sickening | 4 | 25.0 (3.5-177.0) | 94 |
| Purulent secretion in nasal cavity | 3 | 1.8 (1.1-3.2) | 55 | ||
| Purulent rhinorrhea (symptom) | 2 | 0.81 (0.51-1.30) | 35 | ||
| ESR >10 | 0-1 | 0.16 (0.09-0.30) | 10 | ||
| Williams et al,26 1992 | Radiograph | Maxillary toothache | 4-5 | 6.4 (2.2-19.0) | 81 |
| Abnormal transillumination | 3 | 2.6 (1.5-4.4) | 63 | ||
| Poor response to decongestants or antihistamines | 2 | 1.10 (0.73-1.70) | 42 | ||
| 1 | 0.43 (0.27-0.80) | 22 | |||
| Colored nasal discharge Mucopurulence on examination |
0 | 0.16 (0.04-0.41 | 10 | ||
| Van Diujn et al,27 1992 | Ultrasound (A mode) | Beginning with common cold | 5 | … | 89c |
| Purulent rhinorrhea | 1-4 | … | 18-82 | ||
| Pain at bending Unilateral maxillary pain Pain in teeth |
0 | … | 11 | ||
| Diagnosis of ABRS | |||||
| Berg et al,25 1988 | Purulent antral puncture fluid | Purulent rhinorrhea with unilateral predominance | 3-4 | 7.0 (3.9-12.7) | 82 |
| 2 | 1.30 (0.72-1.9) | 46 | |||
| Local pain with unilateral predominance Bilateral purulent rhinorrhea Pus in nasal cavity |
0-1 | 0.06 (0.02-0.17) | 4 | ||
| Ebell et al,36 2017 | Positive bacterial culture from antral fluid | Tender maxillary sinus (unilateral): 2 points | 7-8 | 5.0 (1.8-14) | 77 |
| 4-6 | 1.8 (1.3-2.5) | 55 | |||
| Maxillary toothache: 2 points C-reactive protein >15 mg/L: 2 points Preceding upper respiratory tract illness: 1 point Purulent nasal discharge: 1 point Previous sinusitis episodes: −1 point |
−1-3 | 0.35 (0.21-0.56) | 19 | ||
ABRS = acute bacterial rhinosinusitis; ARS = acute rhinosinusitis; CT = computed tomography; ESR = erythrocyte sedimentation rate; LR = likelihood ratio.
Calculated using the score-specific LR and a pretest probability of 40%, or taken directly from logistic model for each study.
Results obtained from application of a urinalysis dipstick to nasal secretions.
Probability of sinusitis calculated directly from logistic regression, so LR and its CI not estimable. Each variable is assigned a value of 1 for the presence of each finding and 0 for the absence. The score = −2.124 + 1.035*(beginning with common cold) + 0.996*(purulent rhinorrhea) + 0.95*(pain at bending) + 0.64*(unilateral maxillary pain) + 0.606*(pain in the teeth). A person with all findings present would have a maximum score of 2.103. The probability of sinusitis is calculated from exp(score)/[1 + exp(score)], so exp(2.103)/[1 + exp(2.103)]=89%. A person with none of the findings present would have a score of −2.124, or a probability of 11%.
Huang and colleagues used a 4-item score for results of a urine test strip measuring the protein, pH, leukocyte esterase, and nitrite as measures of an inflammatory response in nasal discharge.24 The samples were collected by having patients blow their nose into a piece of plastic wrap, and the scorer assigned either 0 to 2 or 0 to 3 points per item for a total range of 0 to 11 points. Using a reference standard of radiography, the highest strata of scores (≥4) had a LR of 127.24 Van Diujn and colleagues applied a 5-item score based on a logistic regression equation to estimate the probability of sinusitis compared to ultrasonography of the sinuses as the reference standard.27 This study is limited, like all of the others, by lack of prospective validation and the use of an imaging reference standard that is not specific for ABS.
Acute Bacterial Rhinosinusitis
Two studies reported the accuracy of combinations of signs and symptoms for the diagnosis of ABRS (Table 5).25,36 Berg proposed a 4-item score using a reference standard of purulent antral puncture fluid. The score discriminated well: those with a score or 3 or 4 findings present had a LR of 7 for ABRS, while those with ≤1 findings present had a LR of 0.06.25
The most recently reported clinical decision rule used 5 signs or symptoms plus C-reactive protein (see Table 5 for scoring details) to identify patients at low risk (n = 77, 16%), moderate risk (n = 75, 49%), or high risk (n = 23, 73%) for ABRS using positive bacterial culture of antral puncture fluid as the reference standard.36 Unfortunately, none of the above scores have been prospectively validated.
DISCUSSION
Most individual signs and symptoms have limited accuracy for the diagnosis of ARS and ABRS. Purulent secretions seen in the middle meatus and the overall clinical impression are most useful for ruling in ARS when present, while overall clinical impression, absence of a recent or preceding RTI, absence of any nasal discharge, and absence of purulent nasal discharge reduced the likelihood of ARS (Table 3). Since ARS may have a viral etiology, the more important clinical question is how to best diagnose ABRS. We found that the overall clinical impression, cacosmia (an uncommon but highly specific finding), and pain in the teeth were the best predictors of ABRS (Table 4). Other individual signs and symptoms had positive and negative likelihood ratios between 0.5 and 2.0, indicating little diagnostic value.
Clinical decision rules have been developed for a range of other respiratory conditions, including pneumonia,37 streptococcal pharyngitis,38 and influenza.39 We identified 6 clinical rules for the diagnosis of ARS or ABRS, and they have promising accuracy, but none has been prospectively validated. In some cases, including 1 of the clinical decision rules for ABRS, they incorporate point-of-care tests such as C-reactive protein as well as signs or symptoms to increase the accuracy of prediction.37 Another potentially useful point-of-care test is use of a urine dipstick to detect leukocyte esterase or nitrites in nasal discharge (Table 5).24 While it could be incorporated into a clinical decision rule or even used alone, its accuracy requires further validation. While C-reactive protein is accurate and widely used in some countries at the point of care to identify patients more or less likely to have a bacterial respiratory infection,40,41 it is not currently available in most US outpatient settings.11 Finally, a recent systematic review by the author found that older studies of ultrasound found that it to be approximately 80% sensitive for ARS.11 Patients who do not have any sinus fluid detected are therefore at low risk for ARS. Thus, studies that evaluate the ability of modern handheld ultrasound devices to detect sinus fluid are needed.
An important limitation of our findings is the variety of reference standards used to define both ARS and ABRS. Because imaging may detect fluid associated with a viral upper respiratory infection, and cannot distinguish purulent from non-purulent fluid, it is likely to lead to over diagnosis of ABRS. Similarly, some patients with fluid that appears purulent may have a viral infection. Relying on these tests could lead to overtreatment with antibiotics. On the other hand, bacterial culture of antral puncture fluid is likely to be more specific but may lack sensitivity if organisms do not grow in culture. In addition, puncturing the antrum is painful and invasive, making it impractical for use in clinical practice. Ultimately, the true prevalence of ABRS among patients with clinically suspected ARS is likely to be between 31% (prevalence using bacterial culture) and 50% (prevalence using imaging) in adults.
An important question is whether detection of a bacterial pathogen in sinus fluid means that the patient will benefit from antibiotics. Most clinical trials have enrolled patients with clinically suspected sinusitis and found a small benefit (5 additional cures per 100 persons receiving an antibiotic). Trials using imaging (3 radiography, 1 CT) found a larger benefit, suggesting some validity for the concept of imaging as a reference standard.6 To date no randomized trials of antibiotics or other interventions have enrolled patients with ABRS diagnosed by inspection of fluid.
While many studies have attempted to identify combinations of signs or symptoms that diagnose ARS or ABRS, it is equally important to determine which patients are at low likelihood of ABRS (would be unlikely to benefit from an antibiotic), as it is to identify which patients are likely to have an uncomplicated course. Thus, research to determine low risk criteria that help rule out ABRS is needed.
CONCLUSION
Only about one-third of patients with clinically suspected ARS have a positive bacterial culture of antral puncture fluid. Acute rhinosinusitis as diagnosed by any reference standard is significantly less likely in patients without any nasal discharge, without a complaint of purulent nasal discharge, and with normal transillumination. The overall clinical impression is also useful for both ruling in and ruling out ARS. Evidence regarding diagnosis of ABRS is limited, but we conclude that the overall clinical impression, pain in the teeth, and cacosmia are the most useful findings for clinicians trying to identify patients most likely to benefit from antibiotics. Clinical decision rules, including those incorporating C-reactive protein, and the use of urine dipsticks to test the nasal discharge are promising, but all require prospective validation.
Footnotes
Conflicts of interest: authors report none.
Supplementary materials: Available at http://www.AnnFamMed.org/content/17/2/164/suppl/DC1/.
References
- 1.Rosenfeld RM, Piccirillo JF, Chandrasekhar SS, et al. Clinical practice guideline (update): adult sinusitis. Otolaryngology–Head Neck Surg. 2015; 152(2)(Suppl): S1–S39. [DOI] [PubMed] [Google Scholar]
- 2.Sharp HJ, Denman D, Puumala S, Leopold DA. Treatment of acute and chronic rhinosinusitis in the United States, 1999-2002. Arch Otolaryngol Head Neck Surg. 2007; 133(3): 260–265. [DOI] [PubMed] [Google Scholar]
- 3.Fleming-Dutra KE, Hersh AL, Shapiro DJ, et al. Prevalence of inappropriate antibiotic prescriptions among US ambulatory care visits, 2010-2011. JAMA. 2016; 315(17): 1864–1873. [DOI] [PubMed] [Google Scholar]
- 4.Hansen JG, Schmidt H, Rosborg J, Lund E. Predicting acute maxillary sinusitis in a general practice population. BMJ. 1995; 311(6999): 233–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.van Buchem L, Peeters M, Beaumont J, Knottnerus JA. Acute maxillary sinusitis in general practice: the relation between clinical picture and objective findings. Eur J Gen Pract. 1995; 1(4): 155–160. [Google Scholar]
- 6.Lemiengre MB, van Driel ML, Merenstein D, Liira H, Mäkelä M, De Sutter AI. Antibiotics for acute rhinosinusitis in adults. Cochrane Database Syst Rev. 2018; 9(9): CD006089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Varonen H, Mäkelä M, Savolainen S, Läärä E, Hilden J. Comparison of ultrasound, radiography, and clinical examination in the diagnosis of acute maxillary sinusitis: a systematic review. J Clin Epidemiol. 2000; 53(9): 940–948. [DOI] [PubMed] [Google Scholar]
- 8.Engels EA, Terrin N, Barza M, Lau J. Meta-analysis of diagnostic tests for acute sinusitis. J Clin Epidemiol. 2000; 53(8): 852–862. [DOI] [PubMed] [Google Scholar]
- 9.Williams JW, Jr, Simel DL. Does this patient have sinusitis? Diagnosing acute sinusitis by history and physical examination. JAMA. 1993; 270(10): 1242–1246. [PubMed] [Google Scholar]
- 10.Ioannidis JP, Lau J. Technical report: evidence for the diagnosis and treatment of acute uncomplicated sinusitis in children: a systematic overview. Pediatrics. 2001; 108(3): E57. [DOI] [PubMed] [Google Scholar]
- 11.Ebell MH, Guilbault R, Ermias Y. Diagnosis of acute rhinosinusitis in primary care: a systematic review of test accuracy. Br J Gen Pract. 2016; 66(650): e612–632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lindbaek M, Hjortdahl P, Johnsen UL. Use of symptoms, signs, and blood tests to diagnose acute sinus infections in primary care: comparison with computed tomography. Fam Med. 1996; 28(3): 183–188. [PubMed] [Google Scholar]
- 13.McNeill RA. Comparison of the findings on transillumination, x-ray and lavage of the maxillary sinus. J Laryngol Otol. 1963; 77: 1009–1013. [DOI] [PubMed] [Google Scholar]
- 14.Bergstedt HF, Carenfelt C, Lind MG. Facial bone scintigraphy. VI. Practical clinical use in inflammatory disorders of the maxillary sinus. Acta Radiol Diagn (Stockh). 1980; 21(5): 651–656. [DOI] [PubMed] [Google Scholar]
- 15.Berg O, Bergstedt H, Carenfelt C, Lind MG, Perols O. Discrimination of purulent from nonpurulent maxillary sinusitis; clinical and radiographic diagnosis. Ann Otol Rhinol Laryngol. 1981; 90(3 Pt 1): 272–275. [DOI] [PubMed] [Google Scholar]
- 16.Hansen JG, Lund E. The association between paranasal computerized tomography scans and symptoms and signs in a general practice population with acute maxillary sinusitis. APMIS. 2011; 119(1): 44–48. [DOI] [PubMed] [Google Scholar]
- 17.Hansen JG, Højbjerg T, Rosborg J. Symptoms and signs in culture-proven acute maxillary sinusitis in a general practice population. APMIS. 2009; 117(10): 724–729. [DOI] [PubMed] [Google Scholar]
- 18.Van Buchem FL, Peeters MF, Knottnerus JA. Maxillary sinusitis in children. Clin Otolaryngol Allied Sci. 1992; 17(1): 49–53. [DOI] [PubMed] [Google Scholar]
- 19.Whiting PF, Rutjes AW, Westwood ME, et al. ; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011; 155(8): 529–536. [DOI] [PubMed] [Google Scholar]
- 20.Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol. 2003; 56(11): 1129–1135. [DOI] [PubMed] [Google Scholar]
- 21.Axelsson A, Runze U. Symptoms and signs of acute maxillary sinusitis. ORL J Otorhinolaryngol Relat Spec. 1976; 38(5): 298–308. [DOI] [PubMed] [Google Scholar]
- 22.Berg O, Carenfelt C. Etiological diagnosis in sinusitis: ultrasonography as clinical complement. Laryngoscope. 1985; 95(7 Pt 1): 851–853. [PubMed] [Google Scholar]
- 23.Visca A, Castello M, De Filippi C. [Diagnostic considerations on sinusitis in childhood]. Minerva Pediatr. 1995; 47(5): 171–174. [PubMed] [Google Scholar]
- 24.Huang SW, Small PA., Jr Rapid diagnosis of bacterial sinusitis in patients using a simple test of nasal secretions. Allergy Asthma Proc. 2008; 29(6): 640–643. [DOI] [PubMed] [Google Scholar]
- 25.Berg O, Carenfelt C. Analysis of symptoms and clinical signs in the maxillary sinus empyema. Acta Otolaryngol. 1988; 105(3-4): 343–349. [DOI] [PubMed] [Google Scholar]
- 26.Williams JW, Jr, Simel DL, Roberts L, Samsa GP. Clinical evaluation for sinusitis: making the diagnosis by history and physical examination. Ann Intern Med. 1992; 117(9): 705–710. [DOI] [PubMed] [Google Scholar]
- 27.van Duijn NP, Brouwer HJ, Lamberts H. Use of symptoms and signs to diagnose maxillary sinusitis in general practice: comparison with ultrasonography. BMJ. 1992; 305(6855): 684–687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Laine K, Määttä T, Varonen H, Mäkelä M. Diagnosing acute maxillary sinusitis in primary care: a comparison of ultrasound, clinical examination and radiography. Rhinology. 1998; 36(1): 2–6. [PubMed] [Google Scholar]
- 29.Varonen H, Savolainen S, Kunnamo I, Heikkinen R, Revonta M. Acute rhinosinusitis in primary care: a comparison of symptoms, signs, ultrasound, and radiography. Rhinology. 2003; 41(1): 37–43. [PubMed] [Google Scholar]
- 30.Thomas C, Aizin V. Brief report: a red streak in the lateral recess of the oropharynx predicts acute sinusitis. J Gen Intern Med. 2006; 21(9): 986–988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Shaikh N, Hoberman A, Kearney DH, et al. Signs and symptoms that differentiate acute sinusitis from viral upper respiratory tract infection. Pediatr Infect Dis J. 2013; 32(10): 1061–1065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Autio TJ, Koskenkorva T, Närkiö M, Leino TK, Koivunen P, Alho OP. Diagnostic accuracy of history and physical examination in bacterial acute rhinosinusitis. Laryngoscope. 2015; 125(7): 1541–1546. [DOI] [PubMed] [Google Scholar]
- 33.Ebell MH, McKay B, Guilbault R, Ermias Y. Diagnosis of acute rhinosinusitis in primary care: a systematic review of test accuracy. Br J Gen Pract. 2016. September;66(650):e612–32. Epub 2016 Aug 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Puhakka T, Heikkinen T, Mäkelä MJ, et al. Validity of ultrasonography in diagnosis of acute maxillary sinusitis. Arch Otolaryngol Head Neck Surg. 2000; 126(12): 1482–1486. [DOI] [PubMed] [Google Scholar]
- 35.Schwartz K, Monsur J, Northrup J, West P, Neale AV. Pharyngitis clinical prediction rules: effect of interobserver agreement: a MetroNet study. J Clin Epidemiol. 2004; 57(2): 142–146. [DOI] [PubMed] [Google Scholar]
- 36.Ebell MH, Hansen JG. Proposed clinical decision rules to diagnose acute rhinosinusitis among adults in primary care. Ann Fam Med. 2017; 15(4): 347–354. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.van Vugt SF, Broekhuizen BD, Zuithoff NP, et al. ; GRACE Consortium. Validity of a clinical model to predict influenza in patients presenting with symptoms of lower respiratory tract infection in primary care. Fam Pract. 2015; 32(4): 408–414. [DOI] [PubMed] [Google Scholar]
- 38.Centor RM, Witherspoon JM, Dalton HP, Brody CE, Link K. The diagnosis of strep throat in adults in the emergency room. Medical Decis Making. 1981; 1(3): 239–246. [DOI] [PubMed] [Google Scholar]
- 39.Ebell MH, Afonso AM, Gonzales R, Stein J, Genton B, Senn N. Development and validation of a clinical decision rule for the diagnosis of influenza. J Am Board Fam Med. 2012; 25(1): 55–62. [DOI] [PubMed] [Google Scholar]
- 40.Cals JW, Schot MJ, de Jong SA, Dinant GJ, Hopstaken RM. Point-of-care C-reactive protein testing and antibiotic prescribing for respiratory tract infections: a randomized controlled trial. Ann Fam Med. 2010; 8(2): 124–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Llor C, Bjerrum L, Arranz J, et al. C-reactive protein testing in patients with acute rhinosinusitis leads to a reduction in antibiotic use. Fam Pract. 2012; 29(6): 653–658. [DOI] [PubMed] [Google Scholar]