Background

In March 1999, the Agency for Healthcare Research and Quality (previously the Agency for Health Care Policy and Research) published the original Evidence Report/Technology Assessment No. 9, Diagnosis and Treatment of Acute Bacterial Rhinosinusitis, which was prepared by the New England Medical Center Evidence-based Practice Center (EPC) under contract to the Agency. The original report used relevant, high-quality studies, but they dealt mostly with the diagnosis and treatment of acute rhinosinusitis in adults. In preparing that report, the EPC found only two randomized controlled trials that were devoted exclusively to the efficacy of using antibiotics to treat this condition in children.

This supplement was developed by the EPC to identify and analyze evidence from nonrandomized studies that pertains to the diagnosis and therapeutic management of acute sinusitis in children. Compared with adults, children have a different and constantly evolving sinus anatomy, and they probably have an increased incidence of upper respiratory tract infections.

Research Questions

This report deals with uncomplicated acute sinusitis, which is typically defined by symptoms that have persisted for less than 30 days. The following research questions were used to guide the analysis:

1. What is the evidence for the efficacy of various antibiotics in children with a diagnosis of acute sinusitis?
2. What is the evidence for the efficacy of various ancillary regimens that do not include antibiotics in the treatment of children with acute sinusitis?
3. What is the diagnostic accuracy and concordance of clinical symptoms, radiography and other imaging methods, and aspiration for the diagnosis of acute sinusitis in children?

Methodology

Studies were included in this report-regardless of study design-if they were pertinent to the research questions, involved only patients younger than 18 years of age (or included subgroups of patients under age 18 that could be readily identified in the data), included at least 10 children, and involved only children who had been symptomatic for less than 30 days. Studies of chronic sinusitis were excluded, as were studies limited to complications (neurologic, local soft tissue, or other) of acute sinusitis.

To identify relevant studies, the authors of the report searched MEDLINE using a broad search strategy covering the period from 1966 to March 1999. The word "sinusitis" was used in the search as a text word and as a MeSH term. Search results were limited to English-language, human studies that included pediatric patients. The titles and abstracts of 1,857 citations were retrieved and screened; 1,719 articles were rejected immediately on the basis of their titles and abstracts. The remaining 138 articles not examined for the previous report were retrieved in full and examined. These articles included mostly nonrandomized studies and a few new randomized trials. A total of 21 studies ultimately qualified for inclusion in the supplement, including five randomized trials and eight nonrandomized studies on antibiotic therapy. These 13 studies were published between 1970 and 1997 and with two exceptions had been conducted at single centers by pediatricians or otolaryngologists. The largest case series had only 106 patients, and the largest randomized controlled trial had only 93 patients. Overall, 255 children were studied in the five randomized controlled trials, and 418 children were studied in the eight nonrandomized studies.

Results

Eight of the 13 studies of antibiotic therapy did not specify the duration of symptoms, and therefore the diagnosis of "acute sinusitis" was not certain. Puncture for aspiration/irrigation was performed in selected children in six studies. Positive radiographic findings were required for the diagnosis of acute sinusitis in nine of the 13 studies, and no other imaging was performed. Clinical symptoms and signs typically were key factors in diagnosis, but there was substantial variation in the way sinusitis was diagnosed and in the prevalence of specific symptoms and signs.

A large array of antibiotics was tested. A placebo arm was present in two randomized controlled trials, and a "no antibiotic treatment group" was included in one of the case series. The duration of treatment varied from 3 to 28 days. The two shorter courses (3 and 5 days) were with azithromycin; all other studies involved at least 7 days of therapy. Decongestants either were reported to be routinely prescribed, or their use was not mentioned at all. One randomized controlled trial looked at whether lavage provided additional benefit.

Response to treatment was assessed typically after 7 to 14 days and, in some cases, later. With the limited data available, the clinical improvement rate with antibiotics was estimated at 87.6 percent (177/202) in randomized controlled trials and 92.2 percent (318/345) in nonrandomized studies. The rate of improvement without antibiotics was 66 percent (33/50). It was 60 percent (21/35) in one randomized trial and 80 percent (12/15) in one observational study.

Response data were available from five imaging (plain film radiography or ultrasound) studies, with an overall improvement rate of 80.4 percent (303/377). Bacteriological data were too scanty and meaningless to combine, and reporting of safety data was erratic. Discontinuation of treatment because of adverse effects was mentioned only in the two largest randomized trials and in three of the eight nonrandomized studies.

Three trials, with a total of 243 patients, studied the efficacy of ancillary measures in the treatment of acute sinusitis in children. One study enrolled children who had sinusitis on the basis of ultrasonography in the absence of any symptoms and addressed the value of lavage versus no lavage on top of background therapy with amoxicillin and phenylpropanolaminohydrochloride. The pertinence of this study to clinical practice is highly questionable. The other two trials evaluated in a double-blind fashion the efficacy of steroid or combination agents (nasal spray budesonide and a combination of nasal oxymetazoline in addition to oral brompheniramine or phenylpropanolamine, respectively) against placebo. None of the three studies used cure-improvement-failure categorization for clinical outcomes. No significant difference was found at any of the addressed time points, except for a superiority of budesonide over placebo at the end of 2 weeks in terms of the clinical score.

A total of eight studies provided some data on the performance of diagnostic tests. Five of these studies addressed the comparative diagnostic accuracy of at least two diagnostic procedures in acute sinusitis in children. Two of the randomized controlled trials on therapeutic measures provided data on percentage of abnormal radiographs among children with symptoms of sinusitis. One study compared ultrasound and radiography in a subgroup of children found incidentally to have sinusitis by occipitomental radiography when they were hospitalized for adenotonsillectomy or adenoidectomy. Typically, radiologists and/or otolaryngologists, rather than general pediatricians, authored the reports, and the study population usually was not adequately defined in terms of symptom duration.

One study found good correlation between ultrasonographic findings and retrieval of fluid upon aspiration, but cultures of the aspirate from 59 sinuses yielded microbial pathogens in less than half (26/59) of the cases. The only study to compare ultrasonography with plain radiography and sinus fluid abnormalities in children with a clinical picture of sinusitis found very low concordance between these diagnostic techniques. Moreover, nonclear irrigation fluid had no correlation with the presence of pathogenic microorganisms.

Conclusion

This report examined the available evidence from randomized trials and nonrandomized studies on the diagnosis and management of acute sinusitis in children. The major conclusion is that, compared with the frequency of this very common condition, the amount of high-quality evidence is remarkably limited.

There is very little evidence on how to accurately diagnose acute sinusitis in childhood. Plain film radiography shows only modest concordance with clinical diagnosis, and the concordance depends largely on how a clinical diagnosis is defined. Other imaging modalities and irrigation have no clear role in the diagnostic management of the syndrome. There is no consensus on which clinical signs and symptoms are most useful for diagnosing this condition, and very limited attention has been given to this issue. Although one small trial found antibiotics to be superior to placebo, its applicability to settings in which sinusitis is defined by different criteria is uncertain. The available evidence also suggests that the various antibiotics used for pediatric sinusitis do not differ in their efficacy rates. In the absence of a gold standard for diagnosis, trials involving several hundred children would be needed to show such differences. Finally, there is no convincing evidence to support the use of ancillary treatment with decongestant-antihistamines and very limited evidence on the use of steroids.

This investigation clearly documents the paucity of the evidence and identifies important questions that need to be addressed in future studies. The paucity of primary data may be due to the difficulties when studying a pediatric population of applying the necessary rigorous methodologies that are needed to generate high-quality information. Obviously, more evidence-based research on this common infection is needed.

Research Questions

With these differences in mind, the main questions addressed in this report are as follows:

1. What is the evidence for the efficacy of various antibiotics in children with a diagnosis of acute sinusitis?
2. What is the evidence for the efficacy of various ancillary regimens that do not include antibiotics in the treatment of children with acute sinusitis?
3. What is the diagnostic accuracy and concordance of clinical symptoms, radiography and other imaging methods, and aspiration for the diagnosis of acute sinusitis in children?

Definitions

As in the original evidence report, we focused on acute uncomplicated bacterial sinusitis. Inflammation of the paranasal sinuses may be accompanied by inflammation of the nasal passages. As such, the clinical condition termed "sinusitis" may in fact be rhinosinusitis. We excluded cases where clinically evident neurological, soft tissue, or other complications were present. Usually, these cases present as medical or surgical emergencies. Acute sinusitis is defined typically by symptoms that have persisted for less than 30 days. We did not attempt to separate bacterial from nonbacterial cases in the considered reports, as this would have been impossible. But, obviously, determining the effectiveness of antibiotic management is important for the treatment of cases with a bacterial etiology.

Cure, improvement, and treatment failure definitions are based on the original reports. Cures and failures were recorded as defined by the individual study. Cure generally meant resolution of all signs and symptoms, and failure generally signified no change or a worsening of signs and symptoms. Finally, in the absence of a gold standard for diagnosing "acute uncomplicated bacterial rhinosinusitis" (other than sinus aspiration and culture, which is very infrequently used), diagnostic parameters in the comparison of diagnostic methods should be interpreted as estimates of concordance between these diagnostic modalities rather than proof of diagnostic accuracy or lack thereof.

Inclusion Criteria

Studies qualified for inclusion in this report regardless of study design if they

• Were pertinent to the questions identified above.
• Studied only patients younger than 18 years of age (or included subgroups of patients under age 18 that could be readily identified in the presented data).
• Included at least 10 children.
• Included only patients who were symptomatic for less than 30 days.

Studies of chronic sinusitis were excluded. Subgroup data on acute sinusitis in studies with 10 or more patients in which both acute and chronic sinusitis or other infections were considered qualified for inclusion in this evidence report. Studies limited to complications (neurologic, local soft tissue, or other) of acute sinusitis were excluded.

Search Strategy and Retrievals

We searched MEDLINE using a broad search strategy covering the period from 1966 to March 1999. The word "sinusitis" was used in the search as a text word and as a MeSH term. We then limited the search results to human studies and English-language studies that included pediatric patients using the terms "infant, newborn," "infant," "child, preschool," "child," and "adolescent." The titles and abstracts of the resulting 1,857 citations were retrieved and screened for articles that might have data on treatment of acute sinusitis in children. We immediately rejected 1,719 papers on the basis of their titles and abstracts. Notably, these included 450 articles on complications of acute sinusitis and 233 nonsystematic review articles without original data. The remaining 138 articles not previously examined in the earlier evidence report (Lau, Zucker, Engels, et al., 1999) were retrieved in full and examined. These articles included mostly nonrandomized studies and a few new randomized trials. A total of 21 studies ultimately qualified for inclusion in this supplemental report.

Statistical Analysis

Given the paucity and heterogeneity of the data for specific questions, we did not attempt the application of formal meta-analytic techniques in most circumstances. When possible, rates were pooled across different studies, and heterogeneity was assessed with a chi-square statistic. Odds ratios for efficacy (clinical improvement) were also estimated by the Mantel-Haenszel formula stratified per antibiotic use.

For diagnostic modalities, we expressed concordance by using the positive likelihood ratio, which is calculated as

and the negative likelihood ratio, which is calculated as

The positive likelihood ratio gives an estimate of how much more common a specific diagnostic finding is in the positive group versus the negative group as defined by a different diagnostic standard. The higher the positive likelihood ratio, the better the concordance of the two diagnostic modalities. A positive likelihood ratio of 1 shows there is no concordance. The positive likelihood ratio can take values up to infinity. There is no threshold value that characterizes a test of very good diagnostic performance, but values less than 5 probably are not sufficient. For example, a positive likelihood ratio of 5 corresponds to a sensitivity of 50 percent with specificity of 90 percent or a sensitivity of 90 percent with specificity of 82 percent. The inverse considerations hold true for the negative likelihood ratio, where good concordance is shown by diminishing values. A negative likelihood ratio of 1 also shows lack of concordance.

All reported p-values are two-tailed.

Efficacy of Antibiotic Interventions

As shown in Evidence Table 1a, five randomized trials and eight nonrandomized studies qualified for inclusion in this analysis. Of 68 randomized trials on acute sinusitis that we identified as part of the previous sinusitis evidence report (Lau, Zucker, Engels, et al., 1999) and in this supplemental report, only five involved children exclusively. Two of the five randomized trials were used in the meta-analyses of the main evidence report (Wald, Chiponis, and Ledesma-Medina, 1986; Wald, Reilly, Casselbrant, et al., 1984). The remaining three studies (Aitken and Taylor, 1998; Careddu, Bellosta, Tonelli, et al., 1993; Jeppesen and Illum, 1972) did not qualify for inclusion in the original meta-analyses.

In 30 of the 68 randomized trials identified in the earlier report (Lau, Zucker, Engels, et al., 1999), the age range of the enrolled patients extended to less than 18 years, but no separate data on children under 18 years of age were available; most of the patients presumably were adults. Of these 30 trials, 23 had a lower age limit of between 12 and 17 years, and for another five it was 10 or 11 years. For all these 30 excluded studies, the upper age limit was 60 years or undefined, except for one study with an upper limit of 40 years. The remaining 38 randomized trials either included only adults or did not report any age data.

Description of Included Studies

The 13 qualifying studies of antibiotic therapy had been published between 1970 and 1997 and with two exceptions had been conducted at single centers by pediatricians or otolaryngologists. Nine of the 13 reports, including six of the eight nonrandomized studies, originated outside the United States. Pharmaceutical sponsorship was clearly mentioned in four cases, but it is probable that most of these studies were funded by the pharmaceutical industry. The largest case series had only 106 patients, and the largest randomized controlled trial had only 93 patients. Overall, 255 children were studied in the five randomized controlled trials, and 418 children were studied in the eight nonrandomized studies. Eight of the 13 reports did not specify the duration of symptoms, and therefore we could not be certain of the diagnosis "acute sinusitis." Puncture for aspiration/irrigation was performed in selected children in six studies. Positive radiographic findings (typically combinations of air-fluid level, opacification, and/or mucous thickening criteria) were required for the diagnosis of acute sinusitis in nine of the 13 studies; no other imaging was performed in these studies. Clinical symptoms and signs typically were key factors in diagnosis, but there was substantial variation in the way sinusitis was diagnosed and in the prevalence of specific symptoms and signs.

As shown in Evidence tables 1a--1e, a large array of antibiotics was tested. A placebo arm was present in two randomized controlled trials (Jeppesen and Illum, 1972; Wald, Chiponis, and Ledesma-Medina, 1986), and a "no antibiotic treatment group" was included in one of the case series (Aitken and Taylor, 1998). The duration of treatment varied between 3 and 28 days. The two shorter courses (3 and 5 days) were with azithromycin, which results in high drug levels for several days after its discontinuation. All other studies involved at least 7 days of therapy. Either decongestants were reported to be routinely prescribed, or their use was not mentioned at all. One randomized controlled trial looked at whether lavage provided additional benefit (Jeppesen and Illum, 1972).

Response to treatment was assessed typically after 7--14 days and, in some cases, later. Cure, improvement, and failure rates are shown in evidence table 1d. Overall, using the available data, we estimated that the clinical improvement rate with antibiotics was 87.6 percent (177/202) in randomized controlled trials and 92.2 percent (318/345) in nonrandomized studies. The rate of improvement without antibiotics was 66 percent (33/50). It was 60 percent (21/35) in one randomized trial (Wald, Chiponis, and Ledesma-Medina, 1986) and 80 percent (12/15) in one observational study (Aitken and Taylor, 1998). The only randomized trial that compared antibiotics with placebo found significantly better efficacy for antibiotics (Wald, Chiponis, and Ledesma-Medina, 1986). No significant differences were seen in the efficacy of various antibiotic regimens in direct randomized comparisons. Overall, there was a trend of higher improvement rates in the nonrandomized studies compared with the randomized studies (odds ratio, 1.79 [95 percent CI, 1.05-3.04], stratified for antibiotic use). The same trend was found for improvement rates with antibiotics (odds ratio, 1.66) and without antibiotics (odds ratio, 2.67). Improvement rates did not differ significantly among the various individual studies.

Response data were available from five imaging (plain film radiography or ultrasound) studies (Helin, Andreasson, Jannert, et al., 1982; Herz and Gfeller, 1977; McLean, 1970; Revonta and Suonpaa, 1982; Wald, Reilly, Casselbrant, et al., 1984), yielding an overall improvement rate of 80.43 percent (303/377). Bacteriological data were too scanty and meaningless to combine. Reporting of safety data was erratic. Discontinuation of treatment because of adverse effects was mentioned only in the two largest randomized trials (Careddu, Bellosta, Tonelli, et al., 1993; Wald, Chiponis, and Ledesma-Medina, 1986) and in three of the eight nonrandomized studies (Gurses, Kalayci, Islek, et al., 1996; Hager, Bamberg, Dorn, et al., 1980; Herz and Gfeller, 1977).

Efficacy of Ancillary Measures

Of the 12 randomized trials on ancillary measures identified as part of the previous report (Lau, Zucker, Engels, et al., 1999), only three trials (Barlan, Erkan, Bakir, et al., 1997; McCormick, John, Swischuk, et al., 1996; Revonta and Suonpaa, 1982) with a total of 243 patients studied the efficacy of ancillary measures in the treatment of acute sinusitis in children (Evidence tables 2a--2d). The age range of patients in the remaining nine trials extended to as low as 9 to 20 years (>12 years in seven trials). No separate data on children were provided, and most of the enrolled patients presumably were adults (upper age limit of 62 or undefined).

Of the three qualifying trials (Barlan, Erkan, Bakir, et al., 1997; McCormick, John, Swischuk, et al., 1996; Revonta and Suonpaa, 1982), one enrolled children who had sinusitis on the basis of ultrasonography in the absence of any symptoms and addressed the value of lavage versus no lavage on top of background therapy with amoxicillin and phenylpropanolaminohydrochloride (Revonta and Suonpaa, 1982). The pertinence of this study to clinical practice is highly questionable. The other two trials (Barlan, Erkan, Bakir, et al., 1997; McCormick, John, Swischuk, et al., 1996) evaluated in a double-blind fashion the efficacy of steroid or combination agents (nasal spray budenoside and a combination of nasal oxymetazoline in addition to oral brompheniramine or phenylpropanolamine, respectively) against placebo.

None of the three studies used cure-improvement-failure categorization for clinical outcomes. The study on lavage used strictly ultrasonographic criteria (Revonta and Suonpaa, 1982). The other two (Barlan, Erkan, Bakir, et al., 1997; McCormick, John, Swischuk, et al., 1996) used composite clinical and/or radiologic scores. No significant difference was found at any of the addressed time points, except for a superiority of budenoside over placebo at the end of 2 weeks in terms of the clinical score. In the Barlan, Erkan, Bakir, et al. study (1997), only 89 of the 151 enrolled patients received sufficient followup to be included in the analysis.

Accuracy and Concordance of Diagnostic Methods

Eight studies provided some data on the performance of diagnostic tests. Five studies (Glasier, Mallory, and Steele, 1989; Jannert, Andreasson, Helin, et al., 1982; Kogutt and Swischuk, 1973; van Buchem, Peeters, and Knottnerus, 1992; Watt-Boolsen and Karle, 1977) addressed the comparative diagnostic accuracy of at least two diagnostic procedures in acute sinusitis in children (Evidence tables 3a--3d). In addition, two of the randomized controlled trials on therapeutic measures (Barlan, Erkan, Bakir, et al., 1997; Wald, Chiponis, and Ledesma-Medina, 1986) provided data on percentage of abnormal radiographs among children with symptoms of sinusitis. One study (Revonta and Suonpaa, 1982) compared ultrasound and radiography in a subgroup of children found incidentally to have sinusitis by occipitomental radiography when they were hospitalized for adenotonsillectomy or adenoidectomy. These studies were very heterogeneous. Five studies (Barlan, Erkan, Bakir, et al., 1997; Jannert, Andreasson, Helin, et al., 1982; Revonta and Suonpaa, 1982; van Buchem, Peeters, and Knottnerus, 1992; Watt-Boolsen and Karle, 1977) originated outside the United States. Typically, radiologists and/or otolaryngologists, rather than general pediatricians, authored the reports. The study population usually was not adequately defined in terms of symptom duration.

One study addressed a population that had no sinusitis symptoms at all (Revonta and Suonpaa, 1982), but sinusitis was diagnosed by ultrasonography. Plain film radiography was done in six studies (Barlan, Erkan, Bakir, et al., 1997; Jannert, Andreasson, Helin, et al., 1982; Kogutt and Swischuk, 1973; van Buchem, Peeters, and Knotterneus, 1992; Wald, Chiponis, and Ledesma-Medina, 1986; Watt-Boolsen and Karle, 1977) where some or all patients had clinical symptoms of sinusitis. Plain film radiography was considered to be worthless in a study of infants; computed tomography was used instead (Glasier, Mallory, and Steele, 1989). Although radiographs were performed in the Revonta and Suonpaa (1982) study, it was unclear whether this study population had sinusitis symptoms. In the six studies in which plain film radiography was done, the rate of abnormal findings (typically opacification, mucosal thickening, and/or air-fluid level) against a clinical diagnosis of sinusitis ranged from 55.4 percent to 96 percent. The pooled rate was 73.2 percent (596/814), and heterogeneity between studies was high (p < 0.001). Most of this variability is probably due to differences in the clinical definition of sinusitis, which could be discerned easily in the only study (Jannert, Andreasson, Helin, et al., 1982) that used different thresholds for the clinical definition. When the subgroup of children who had only one of the three criteria-purulent secretion, history of upper respiratory secretion, and sinus pain or tenderness-were considered, radiographic abnormalities were present only in 22/79 cases (27.8 percent). When two or three of these criteria were present, radiographic abnormalities were noted in 75/96 cases (78.1 percent). This gives a positive likelihood ratio of 2.81 for radiographic examination to be abnormal in children with more than one clinical criterion compared with children who have only one of the clinical criteria (Evidence table 4).

Similar rates of abnormal radiographs were also seen in the two randomized controlled trials that used strict clinical criteria for the diagnosis of acute sinusitis. Wald, Chiponis, and Ledesma-Medina (1986) defined sinusitis by the presence of any nasal discharge and/or cough that was not improving for 10--30 days. Barlan, Erkan, Bakir, et al. (1997) defined sinusitis by the presence of at least two of three major criteria (purulent nasal discharge, pharyngeal discharge, cough) or one of these plus two of nine minor criteria with a duration of at least 7 days. In these two randomized controlled trials, abnormal radiographs were seen in 136/171 (79.5 percent) and 69/89 (77.5 percent) of children with a clinical diagnosis, respectively. The other three studies that present data on radiography and clinical diagnosis do not specify a priori explicit criteria for the clinical diagnosis of acute sinusitis (Kogutt and Swischuk, 1973; van Buchem, Peeters, and Knottnerus, 1992; Watt-Boolsen and Karle, 1977). One study simply lists the percentage of various symptoms (Kogutt and Swischuk, 1973), while another (Watt-Boolsen and Karle, 1977) does not give any clinical information on signs and symptoms. The third study (van Buchem, Peeters, and Knottnerus, 1992) states that the distinction between "sinusitis" and "rhinitis" was based on the impression of the clinician. Interestingly, the "rhinitis" group did not differ from the "sinusitis" group in the prevalence of fever, purulent secretion, sinus tenderness, or headache. The only differences between the two groups were in cough (54/68 [79.4 percent] vs. 39/79 [49.4 percent]) and prolonged runny nose (72 percent vs. 36 percent), but these differences were not impressive.

In a study of infants, plain radiographs were considered worthless, and thus only the use of computed tomography (CT) scan was evaluated (Glasier, Mallory, and Steele, 1989). Excluding cases of sinus hypoplasia, evidence of CT involvement of the maxillary sinus(es) had an 86.7 percent (13/15) sensitivity but only 40.6 percent (28/69) specificity against the clinical impression of upper respiratory infection symptoms. The positive predictive value is only 13/54 (24.1 percent), and the negative predictive value is 28/30 (93.3 percent). The respective figures for the ethmoid sinus(es) were sensitivity, 66.7 percent (10/15); specificity, 61.3 percent (46/75); positive predictive value, 10/39 (25.6 percent); and negative predictive value, 46/51 (91.0 percent). The concordance of CT scan and clinical impression in infants is thus very poor (Evidence table 4).

One study found good correlation between ultrasonographic findings and retrieval of fluid upon aspiration: 68 of 72 sinuses with ultrasonographic abnormalities yielded fluid upon aspiration (Revonta and Suonpaa, 1982). Cultures of the aspirate from 59 sinuses yielded microbial pathogens in less than half (26/59) of the cases. The only study to compare ultrasonography with plain radiography and sinus fluid abnormalities among children with a clinical picture of sinusitis (van Buchem, Peeters, and Knottnerus, 1992) found very low concordance between these diagnostic techniques (Evidence table 4). Moreover, nonclear irrigation fluid had no correlation with the presence of pathogenic microorganisms. Finally, radiologic abnormalities showed poor concordance even with the presence of fluid on aspiration in one study (Evidence table 4).