Variability in Physical Exam Documentation for Obstructive Sleep Apnea

ABSTRACT

Objectives

This study evaluates variability in obstructive sleep apnea (OSA) physical exam templates, contrasts documentation between institutional settings, and identifies inter‐specialty differences in documentation.

Methods

Cross‐sectional analysis was performed using the Epic Community Library, a shared database of user‐created documentation templates from institutions using the Epic electronic medical record system. Physical exam templates assessing sleep apnea were identified and evaluated against a checklist of 25 key physical exam characteristics from a clinical guideline.

Results

Of 6621 templates from 478 institutions, 579 relevant templates were identified. The most frequently included features were vital signs (89.1%), tonsil size (69.5%), and septal deviation (60.2%). The least frequently recorded features were waist‐hip ratio (0.8%), hyomental distance (1.6%), and Cottle maneuver (1.6%). The median template included only 7 out of our 25 key features, with an IQR of 5. Body mass index (BMI) was documented in 36.7% of exams. Of the templates from academic centers, 63% were unique compared to 14% from community centers and 27% from health networks. Otolaryngology templates documented nasal cavity and nasopharyngeal features and specialized airway techniques significantly more frequently than sleep medicine, pulmonology, and unspecified templates. In contrast, sleep medicine and pulmonology documented neck circumference more often than otolaryngology, and sleep medicine recorded Mallampati scores more frequently.

Conclusion

Significant variability exists in OSA physical exam documentation across institutions and specialties. This may contribute to disparities in patient evaluation and treatment. Standardized OSA physical exam templates could provide a more consistent and comprehensive anatomic exam, facilitate subspecialty communication, and support more accurate patient treatment selections.

Level of Evidence

Level 3.

This study demonstrates significant variability in obstructive sleep apnea (OSA) physical exam templates, documentation between institutional settings, and inter‐specialty differences in documentation. This may contribute to disparities in patient evaluation and treatment.

1. Introduction

Obstructive Sleep Apnea (OSA) affects between 9% and 38% of the population and occurs when chronic upper airway collapse during sleep leads to oxygen desaturation and arousal [1]. OSA is associated with adverse long‐term outcomes, including cardiovascular and metabolic disease, depression, and sleep‐related accidents [2, 3, 4, 5]. Due to these severe comorbidities, accurate diagnosis of OSA is vital, and physical exam findings related to airway anatomy are critical [6].

Upper aerodigestive tract characteristics determine the propensity for collapse and obstruction while sleeping [7]. Although polysomnography is the gold standard for diagnosing OSA, it does not specify the anatomic level of obstruction [2]. Identifying the level of anatomic obstruction is an important early step in evaluating a patient for further diagnostic testing and treatment [8]. This is especially crucial when considering surgical intervention, as anatomical relationships significantly affect the success of uvulopalatopharyngoplasty (UPPP) [9, 10]. However, OSA physical examination lacks standardization, and the extent to which key anatomic features are documented—and how this varies across institutions and specialties—is largely unknown [8].

Continuous positive airway pressure (CPAP) is widely considered first‐line treatment for OSA, but its efficacy is limited due to poor patient adherence [11]. A recent study demonstrated higher rates of symptomatic improvement and comparable therapeutic efficacy in patients with a mean baseline AHI of 43 ± 26 who received multilevel upper airway surgery compared to patients treated with CPAP [12]. Insufficient anatomic delineation of the level of obstruction is commonly observed in sleep surgery and the literature, resulting in poor patient selection for targeted anatomic surgery [8, 13]. The success rate of surgical interventions could be increased by consistent evidence‐based physical exams and improved provider communication.

While individual features of physical examination are significantly associated with OSA severity, a comprehensive and evidence‐based physical exam is absent in the medical literature [14, 15, 16]. Established diagnostic tools combine patient history, brief examination, and an OSA questionnaire, but are inconsistent, rarely locate the level of obstruction, and often include few physical findings [9, 17, 18, 19]. Poor or inconsistent documentation has been demonstrated to lead to communication errors, adverse events, and increased cost [20].

Variation of OSA physical exam documentation across institutions and specialties has not been established in the literature. This is particularly relevant as multiple specialties, including otolaryngology, are involved in diagnosing and managing OSA. Differences in referral patterns and treatment approaches have been reported, with OSA patients treated by otolaryngology being younger, having lower BMI and neck circumference, and having fewer comorbidities [21]. Additionally, sleep specialists employed different treatments compared to primary care, with one‐third of patients referred to specialists receiving an alternative therapy to CPAP [22]. A comprehensive physical exam is crucial for assessing diverse clinical presentations of OSA and could lead to more consistent documentation and precise treatment strategies across various specialties and clinical settings.

The electronic medical record (EMR) provides an accessible and practical means of evaluating physical exam consistency because it is commonly used to standardize procedures, such as physical exams [23]. A structured, evidence‐based physical examination is likely to be helpful for pre‐test diagnostic risk profiling and decision‐making regarding treatment modalities. The purpose of this study is to evaluate the variability in documented OSA physical exam features, compare documentation practices between institutional settings, and identify inter‐specialty differences in documentation.

2. Materials and Methods

This study assessed OSA physical exam characteristics through a cross‐sectional analysis of the Epic Community Library. This database contains user‐generated exam templates created and deployed at institutions that use the Epic EMR. These documents use the SmartText format, which allows users to create and insert predefined text blocks and provides a means of standardizing documentation for common exams.

Each institution in the Community Library was queried using the search terms “obstructive sleep apnea physical exam” and “sleep apnea” between February 2024 and July 2024. Each search result was analyzed by one of the authors to determine if it included a physical exam for sleep apnea. Only templates that contained a physical exam specifically for sleep apnea and were not intended for remote encounters were included. Inter‐institutional duplicate templates, defined as identical templates identified at multiple institutions, were common and recorded during data collection. Inter‐institutional duplicates were only included once in data analysis to select for unique templates, as duplicates would inflate certain exam characteristics and obscure actual variation.

Additional information, such as the institution name, specialty, date of the document, and setting, was recorded for each template. The setting was described as academic for university‐affiliated teaching hospitals, community for other non‐university affiliated centers, and health network for organizations managing multiple institutions. A checklist with 25 key physical exam characteristics was used to evaluate each template found in the Community Library. Because a standardized and comprehensive recommendation for physical exam documentation does not exist for OSA, the checklist was based on physical exam suggestions from a clinical guideline produced by The Adult OSA Task Force of the American Academy of Sleep Medicine (AASM) [24]. To improve the applicability and relevance of the checklist, the authors supplemented it with additional characteristics frequently encountered in clinical practice and the literature. The final checklist was designed to evaluate both guideline‐supported and commonly documented physical exam features to comprehensively assess upper airway anatomy and body habitus‐related factors. The 25 physical exam characteristics on the checklist were organized by anatomic level of airway obstruction.

Each physical exam template was evaluated using the checklist, and inclusion rates for each characteristic on the checklist were calculated. Results were then stratified by institution and by provider specialty. The “N‐1” Chi‐squared test and Student's t‐test were used to assess inter‐institutional and inter‐specialty differences. Statistical testing was performed with Microsoft Excel for Mac (v16.87, Redmond, Washington). Because no human subjects or History of Present Illness were included, Institutional Review Board approval was not required for this study

3. Results

Out of 6621 screened templates from 478 institutions, 579 total templates met the inclusion criteria, of which 128 (22.1%) were unique templates. Of the 478 institutions reviewed, 381 (79.7%) had at least one physical exam template that met the inclusion criteria, and 94 (19.7%) had a unique template. In total, 128 unique templates from 94 institutions were analyzed. The two most common inter‐institutional duplicates were “ENT SLEEP APNEA” and “SLE EXPRESS LANE SLEEP CONSULT”, which respectively comprised 306 (52.8%) and 138 (23.8%) of the total templates found. The dates of each template ranged from 2014 to November 2023, with the majority (89.8%) dated in 2022 or 2023. The search algorithm is shown in Figure 1.

FIGURE 1.

Flow diagram of database review methodology.

Inclusion rates for key physical exam characteristics sorted by anatomy are shown in Figure 2. Overall, the median template included only 7 out of our 25 key features, with an IQR of 5, indicating substantial variability in what is documented. The most frequently documented characteristics were vital signs (89.1%), tonsil size (69.5%), and septal deviation (60.2%). The least frequently documented characteristics that were documented at least once were waist‐hip ratio (0.8%), hyomental distance (1.6%), and Cottle maneuver (1.6%). Thyromental and cricomental distances were not documented at all. Strikingly, BMI and neck circumference were documented in only 36.7% and 39.8% of templates, respectively.

FIGURE 2.

Rates of inclusion of physical exam characteristics on unique templates (n = 128).

Qualitatively, there was a high degree of variation in the physical exam characteristics included across templates. Although most exam templates were formatted by anatomic area, few were comprehensive in their scope, with some exams omitting anatomical areas such as the nasal cavity.

The proportion of unique templates found at each institutional setting (academic, community, and health network) is shown in Figure 3. Of the 379 institutions with a physical exam that fell into these categories, academic centers made up 30.6%, community centers made up 30.6%, and health networks made up 38.8%. When analyzing the proportion of unique exams, 63% of templates from academic centers were unique, while the remainder were inter‐institutional duplicates. This is compared to 14% unique templates from community centers and 27% from health networks. Academic institutions accounted for the most unique exam templates at 57% (n = 128). The difference in the proportion of unique exam templates between academic and community institutions was significant (p = < 0.0001), as well as between academic and health network centers (p = < 0.0001). The difference in proportion of unique templates between community and health network settings was also significant (p = 0.0107).

FIGURE 3.

Comparison of the proportion of unique physical exam templates by institution type (Academic, Community, and Health Network) (n = 379).

The specialties with the most physical exam templates were Sleep Medicine (37.5%), Otolaryngology (18.8%), and Pulmonology (12.5%), as shown in Figure 4. An unspecified category was used to account for the 24.2% of unique physical exam templates for which a specialty could not be identified. The remaining 7% of exams were created either by Neurology, Pediatrics, or Oral and Maxillofacial Surgery (OMS).

FIGURE 4.

Distribution of specialties with a unique physical exam (n = 128).

Otolaryngology physical exam characteristic inclusion rates were compared with the other most‐represented specialties (sleep medicine, pulmonology, and unspecified) and are shown in Table 1. When compared with otolaryngology, sleep medicine documented neck circumference (58% vs. 17%, p = 0.0010) and Mallampati scores (71% vs. 33%, p = 0.0022) more often than otolaryngology, which documented more on nasal features: septal deviation (88% vs. 58%, p = 0.0106), nasal polyps (67% vs. 21%, p = 0.0001), turbinate size (79% vs. 29%, p = 0.0001), nasal congestion (63% vs. 29%, p = 0.0059), adenoid size (50% vs. 2%, p < 0.0001), and Cottle maneuver (8% vs. 0%, p = 0.0486) as well as malocclusion class (46% vs. 17%, p = 0.0092), thyroid size (83% vs. 54%, p = 0.0166), flex laryngoscopy (54% vs. 0%, p < 0.0001), and modified Mueller maneuver (46% vs. 0%, p < 0.0001).

TABLE 1.

Comparing Otolaryngology (n = 24) physical exam characteristic inclusion rate with sleep medicine (n = 48), pulmonology (n = 16), and unspecified (n = 31).

Category	Characteristic	Otolaryngology	Sleep medicine	p	Pulmonology	p	Unspecified	p
General	Vital signs	79%	94%	0.0569	84%	0.6366	100%	0.0529
	BMI	38%	42%	0.7464	42%	0.7663	25%	0.3969
	Neck circumference	17%	58%	0.0010	42%	0.0490	19%	0.8729
	Waist‐hip ratio	0%	2%	0.4885	0%	NA	0%	NA
Oral cavity and oropharynx	Tongue size	42%	56%	0.2659	35%	0.5993	50%	0.6229
	Tonsil size	88%	71%	0.1104	65%	0.0528	63%	0.0648
	Uvula size	38%	33%	0.6764	39%	0.9403	50%	0.4581
	Narrow palate	42%	40%	0.8715	32%	0.4487	25%	0.2762
	Mallampati score	33%	71%	0.0022	52%	0.1627	31%	0.8958
	Friedman's tongue position	8%	6%	0.7501	10%	0.8002	0%	0.2522
Teeth and jaw	Retrognathia	25%	33%	0.4894	35%	0.4292	31%	0.6807
	Overjet	4%	21%	0.0609	19%	0.0978	19%	0.1261
	Malocclusion class	46%	17%	0.0092	6%	0.0006	6%	0.0076
Nasal cavity and nasopharynx	Septal deviation	88%	58%	0.0106	52%	0.0050	50%	0.0090
	Nasal polyps	67%	21%	0.0001	35%	0.0197	31%	0.0274
	Turbinate size	79%	29%	0.0001	35%	0.0013	31%	0.0027
	Nasal congestion	63%	29%	0.0059	42%	0.1258	38%	0.1254
	Adenoid size	50%	2%	< 0.0001	0%	< 0.0001	0%	0.0008
	Cottle maneuver	8%	0%	0.0486	0%	0.1122	0%	0.2522
Neck	Hyomental distance	0%	0%	NA	0%	NA	13%	0.0732
	Thyromental distance	0%	0%	NA	0%	NA	0%	NA
	Cricomental distance	0%	0%	NA	0%	NA	0%	NA
	Thyroid size	83%	54%	0.0166	55%	0.0297	63%	0.1576
Specialized airway	Flex laryngoscopy	54%	0%	< 0.0001	0%	< 0.0001	0%	0.0004
	Modified Mueller maneuver	46%	0%	< 0.0001	0%	< 0.0001	0%	0.0016

Note: p values < 0.05 are shown in bold.

Pulmonology documented neck circumference (42% vs. 17%, p = 0.0490) more often than otolaryngology, which documented the following features more: septal deviation (88% vs. 52%, p = 0.0050), nasal polyps (67% vs. 35%, p = 0.0197), turbinate size (79% vs. 35%, p = 0.0013), adenoid size (50% vs. 0%, p < 0.0001), malocclusion class (46% vs. 6%, p = 0.0006), thyroid size (83% vs. 55%, p = 0.0297), flex laryngoscopy (54% vs. 0%, p < 0.0001), and modified Mueller maneuver (46% vs. 0%, p < 0.0001).

Otolaryngology reported the following features more frequently when compared to the unspecified templates: septal deviation (88% vs. 50%, p = 0.0090), nasal polyps (67% vs. 31%, p = 0.0274), turbinate size (79% vs. 31%, p = 0.0027), adenoid size (50% vs. 0%, p = 0.0008), malocclusion class (46% vs. 6%, p = 0.0076), flex laryngoscopy (54% vs. 0%, p = 0.0004), and modified Mueller maneuver (46% vs. 0%, p = 0.0016).

Overall, otolaryngology documented more nasal cavity and nasopharyngeal features than each of the three comparison specialties. Additionally, specialized airway techniques and malocclusion classes were more frequently included in otolaryngology templates than in the other groups.

4. Discussion

Despite its significant morbidity, OSA is underdiagnosed, and the literature lacks a comprehensive and standardized physical exam [25]. This study demonstrates significant variability and lack of consistency in physical exam templates found within a widely used EMR. Differences in the number of unique physical exam templates were seen between setting types, which could contribute to disparities in documentation. Additionally, comparison of the most highly represented specialties showed significant inconsistencies in the anatomic regions emphasized in documentation. These differences may contribute to variation in diagnostic approach, ultimately affecting the treatment patients receive. All these issues may combine to diminish effective communication between healthcare providers, contributing to poor OSA treatment outcomes.

Adequate and appropriate physical examination is an effective way to prevent medical error and adverse clinical events. In one study of cases involving insufficient physical examination, diagnosis was delayed or missed in 76% of cases, and unnecessary treatment was provided in 18% of cases [20]. The relevant physical exam was not performed in 63% of these cases [20]. Out of the 25 characteristics included in this study, the median template included 7 out of our 25 key features, with an IQR of 5, and only five characteristics were documented more than 50% of the time: vital signs, tonsil size, Mallampati score, septal deviation, and thyroid size. While it is not possible to determine how the templates are used clinically from this data, the significant variability identified in this study may indicate that insufficient OSA physical examination documentation is common.

One significantly under‐documented physical exam feature was BMI, which was recorded in only 36.7% of unique templates. Documenting this characteristic is important considering that obesity is estimated to be the primary driver in 58% of those with moderate or worse sleep‐disordered breathing (apnea‐hypopnea index ≥ 15) [26, 27]. This is due to narrowing of the upper airway because of excess fat deposition, which leads to a greater propensity for collapse during sleep [28]. Similarly, neck circumference, another independent predictor and potentially stronger correlate with OSA severity than BMI, was documented in just 39.8% of templates [29, 30]. Inclusion of BMI and neck circumference in a physical exam ensures that obesity‐related factors are documented and considered when diagnosing and treating OSA. The low rate of BMI and neck circumference documentation we found is consistent with other studies that demonstrate poor documentation and diagnosis of obesity [31, 32]. In one study, just under a third of obese patients who had visited an outpatient setting had a documented diagnosis of obesity [33]. Thorough documentation of body habitus metrics can help identify patients at risk for conditions such as hypertension, cardiovascular disease, and type 2 diabetes, and has the potential to improve early management of OSA and metabolic comorbidities.

In addition to neck circumference, Mallampati scores were more commonly included in sleep medicine and pulmonology templates. For every one‐point increase in Mallampati score, there was an increased odds of OSA (OR 2.5, p = 0.01) and an increase of AHI by 5.2 (p = 0.04) [34]. This is critical as focusing only on specialty‐specific physical exam characteristics in clinic may cause other common indicators of OSA, such as enlarged neck circumference, to be overlooked, leading to incomplete treatment of the disease.

Differences in the rate of unique physical exams between academic, community, and health network settings suggest that academic institutions may be more likely to develop and implement unique physical exam templates compared to community centers. While studies have demonstrated that patients at academic centers tend to have better outcomes [35, 36, 37], the impact of academic medical centers has been shown to positively affect local community centers as well. The presence of an academic medical center within a healthcare market was associated with reduced mortality and more healthy days at home [38]. This suggests that part of the role of an academic center is to provide leadership in diagnostic and treatment approaches. Evidence‐based and thorough physical exam templates can be a part of that impact and implemented as a tool to improve healthcare across setting types.

Inter‐specialty comparison of documented physical exam characteristics demonstrated variation in the anatomic areas observed. Otolaryngology templates documented nasal cavity, nasopharyngeal features, and specialized airway techniques significantly more frequently than the other specialties analyzed. Nasal obstruction has been identified as an independent risk factor for OSA [39, 40]. Using the Starling resistor model to approximate the upper airway, obstruction in the nasal area can lead to suction and decreased intraluminal pressure in the oropharynx, leading to airway collapse in predisposed patients [41, 42]. Induced nasal occlusion and increased degree of nasal occlusion have demonstrated significantly increased rates and duration of apneas and hypopneas [43, 44]. Identifying nasal cavity or nasopharyngeal obstruction is an important step in OSA evaluation that can impact treatment and patient outcomes. Pharmaceutical treatment of nasal obstruction has been demonstrated to reduce apnea‐hypopnea index (AHI) in OSA patients but has a limited effect on sleepiness and sleep architecture [42]. Surgical treatment for nasal occlusion in OSA patients resulted in a mild improvement in AHI (31.6 to 28.9) but significantly improved daytime sleepiness and sleep quality [45]. Reduced nasal obstruction can benefit other methods of treating OSA as well. Improved CPAP adherence was observed in patients with a greater minimum cross‐sectional inferior turbinate area and those with lower nasal resistance [46, 47].

The underrepresentation of maxillofacial exam components in the analyzed templates reveals a significant documentation gap. Key elements of this exam focus on the underlying skeletal formations, assessing for retrognathia, overjet, narrow palate, and malocclusion [48]. Although these findings could signal underlying skeletal constriction that may limit airway volume and contribute to OSA, each of these features was documented in less than 40% of physical exams. Of these metrics, only malocclusion was documented significantly more by one specialty, otolaryngology. Clinically, these findings are essential for identifying patients who may be candidates for highly effective skeletal surgeries, which directly expand the upper airway, such as maxillomandibular advancement [49]. A lack of standardized documentation of these features could lead to an under‐referral of patients who might benefit more from skeletal interventions than from more common soft tissue procedures and CPAP.

Otolaryngology‐specific maneuvers, such as with flexible laryngoscopy, would inherently lead to variation in documentation patterns across specialties, particularly of the upper airway. The fact that these differences emerged as statistically significant in our analysis supports the validity of the dataset in reflecting clinical practice patterns. While other specialists may not be able to comprehensively assess the nasal cavity and nasopharyngeal area without the advanced maneuvers available to otolaryngologists, obstruction at this site is important to consider when diagnosing a patient with OSA, and the appropriate referrals should be made.

Physical examination of upper airway anatomy is important in assessing suitability for UPPP, which is an effective surgery for OSA when obstruction occurs at the level of the soft palate [24, 50]. Key physical exam features to evaluate include the oropharyngeal and palatal structures. Grade 3 tonsillar hypertrophy and Friedman stage I are important predictors of UPPP success, with success rates of 80% and 80.6%, respectively [51, 52]. Friedman staging considers tongue size, soft palate position, tonsil size, and BMI [53]. Additionally, a small neck circumference was found to be a strong predictor of patient satisfaction after UPPP, whereas BMI was not [54]. A comprehensive physical exam is relatively simple to perform and provides important clinical insights into predicting surgical treatment outcomes.

Certain aspects of OSA evaluation may not be represented in this study because they are subjective or procedural rather than part of a standard physical exam. Drug‐induced sleep endoscopy (DISE), for example, is a sedated endoscopic procedure primarily used for airway assessment and surgical planning, and its findings are typically documented in procedure notes rather than exam templates. Similarly, patient‐reported outcome measures such as the Epworth Sleepiness Scale (ESS), the NOSE questionnaire, and the Fatigue Severity Scale (FSS) are usually recorded in the subjective portion of notes and are rarely incorporated into physical exam documentation.

Additionally, this dataset primarily contained templates designed for assessing OSA in adults. In pediatric populations, OSA presents differently, and the physical exam focuses on other elements. Adenoid and tonsillar hypertrophy play a central role in the pathophysiology of pediatric OSA and are given greater emphasis during examination in this age group [19, 55].

This study has several limitations. First, the analysis was limited to templated physical exams from the Epic Community Library, which could introduce selection bias and may not fully represent clinical practice across all institutions and specialties. In 2024, Epic accounted for approximately 42.3% of US hospitals and 54.9% of beds in the EMR market [56]. Its representation is particularly pronounced in academic medical centers, where market share is estimated at 60% or higher [57]. Given Epic's widespread use, this study captures a substantial portion of institutional practice patterns and could provide insight into documentation practices more broadly.

Second, because the study focused on templates rather than their application, it could not determine the frequency of clinical use or whether documented items were routinely performed. While inter‐institutional duplicates were removed to focus on variability, this may underrepresent commonly adopted templates. The 25‐item checklist, while based on a clinical guideline, may not have captured all relevant exam components. Finally, specialty classification relied on user‐provided labels and may be subject to misclassification. Despite these limitations, this large, multi‐institutional dataset offers valuable insight into documentation variability and specialty‐specific trends.

5. Conclusion

Our study reveals significant variability in OSA physical exam documentation across both institutions and specialties and shows that key features like BMI and neck circumference are frequently under‐documented, possibly impacting patient evaluation and treatment strategy. Academic centers had the highest rates of unique templates, illustrating their potential leadership role in healthcare markets. Differences between specialties, such as focus on specific anatomical areas and exam characteristics, demonstrate varied approaches to OSA assessment. Given the involvement of many specialties and institutions in diagnosing OSA, consistent communication through comprehensive documentation is essential for effective and collaborative disease management. Moving forward, using anatomically focused, standardized physical exam templates could encourage targeted assessment and efficient documentation, ultimately leading to better patient outcomes in OSA.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Hauser J., Vacaru A., Punjabi N., and Inman J. C., “Variability in Physical Exam Documentation for Obstructive Sleep Apnea,” Laryngoscope Investigative Otolaryngology 11, no. 1 (2026): e70338, 10.1002/lio2.70338.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.