Comparison of physical examination, laboratory and radiologic findings in the diagnosis of acute appendicitis

Abstract

OBJECTIVE

The aim of the present study was to compare the diagnostic performance of physical examination, routinely used laboratory markers, and radiological imaging modalities in adult patients with suspected acute appendicitis (AA), and to evaluate the relative contribution of each diagnostic tool under real-world clinical conditions.

METHODS

This prospective observational study included 665 consecutive adult patients who underwent appendectomy for AA. Preoperative laboratory parameters, including white blood cell count, C-reactive protein, neutrophil count, neutrophil percentage, total bilirubin, and indirect bilirubin, were recorded. Radiological findings from computed tomography and ultrasonography were documented. Physical examinations were independently performed without access to laboratory or imaging results by both a general surgery resident and a specialist general surgeon. Physical examination findings were categorized as compatible, suspicious, or incompatible with AA. Histopathological examination of all specimens served as the reference standard.

RESULTS

Of the 665 patients, 242 (36.4%) were women, and 423 (63.6%) were men, with a mean age of 34.9±14.2 years. The negative appendectomy rate was 5.8%. Resident physical examination demonstrated the highest sensitivity (95.2%) and overall diagnostic accuracy (88.9%). Among laboratory markers, neutrophil count showed high sensitivity (87.9%), while total and indirect bilirubin demonstrated the highest specificity (89.7% and 94.7%, respectively). Computed tomography achieved a sensitivity of 85.7%, whereas ultrasonography showed lower diagnostic performance, particularly in specificity.

CONCLUSION

Physical examination remains a highly effective diagnostic tool for AA when performed systematically and should be integrated with laboratory and selective imaging rather than replaced by them.

Highlight key points

• Physical examination performed by surgical residents showed the highest diagnostic sensitivity among all evaluated diagnostic tools in acute appendicitis.

• Laboratory inflammatory markers demonstrated heterogeneous diagnostic performance and were insufficient for standalone diagnosis.

• Total and indirect bilirubin levels yielded high specificity, indicating an association with advanced inflammatory appendicitis.

• Computed tomography provided high sensitivity but limited specificity under routine clinical practice conditions.

• Diagnostic accuracy was optimized when physical examination was combined with laboratory parameters and selective imaging rather than used in isolation.

Acute appendicitis (AA) is one of the most common causes of emergency abdominal surgery worldwide and remains a major contributor to emergency surgical admissions in both adult and pediatric populations [1]. Despite advances in medical science & technology, delayed or inaccurate diagnosis of AA may still lead to complications such as perforation, intra-abdominal abscess, and increased morbidity [2].

Although AA was first described more than four centuries ago, its diagnosis continues to pose challenges, particularly in patients with atypical clinical presentations [3]. Over time, technological advances and increased accessibility to diagnostic modalities have substantially altered clinical practice; however, these changes have not consistently resulted in proportional improvements in diagnostic accuracy or clinical outcomes [4]. Consequently, the widespread use of advanced diagnostic tools has raised concerns regarding their potential overuse and limited incremental benefit in routine practice [5].

Clinical evaluation, including history taking and physical examination (PE), remains the cornerstone of AA diagnosis and often guides initial decision-making [6]. Classical symptoms and physical findings continue to play a critical role, yet the diagnostic performance of PE may vary depending on physician experience and patient-related factors [7]. The subjective and non-quantitative nature of PE has contributed to its gradual de-emphasis in favor of laboratory and radiological investigations in contemporary practice [8].

Laboratory parameters such as white blood cell count (WBC) and C-reactive protein (CRP) are widely used adjuncts in the diagnostic evaluation of AA and reflect the systemic inflammatory response [9]. Numerous studies have investigated additional inflammatory markers, including neutrophil count (NEU), neutrophil percentage (NEU%), and bilirubin levels, in an attempt to improve diagnostic accuracy [10]. However, current evidence indicates that no single laboratory marker demonstrates sufficient sensitivity and specificity to independently confirm or exclude AA [11].

Radiological imaging has become increasingly integrated into diagnostic algorithms for suspected AA, with ultrasonography (US) and computed tomography (CT) being the most frequently utilized modalities [12]. While CT has been shown to provide high diagnostic accuracy and reduce negative appendectomy rates, concerns regarding radiation exposure, cost, and resource utilization remain significant limitations [13]. Albeit being radiation-free and widely accessible, US is inherently operator dependent and may demonstrate variable diagnostic performance across institutions [14].

Despite technological progress, several contemporary studies emphasize that excessive reliance on imaging does not necessarily improve diagnostic outcomes and may undermine the value of clinical judgment [15]. Therefore, systematic comparison of PE, laboratory parameters, and radiological modalities remains essential to optimize diagnostic strategies for AA.

Accordingly, the present study aimed to compare the diagnostic performance of PE, routinely used laboratory markers including WBC, CRP, NEU, NEU%, total bilirubin (TBIL), and indirect bilirubin (IBIL), and imaging modalities in terms of US and CT in patients with suspected AA, in order to evaluate the effectiveness of routine diagnostic approaches and identify the relative advantages of each diagnostic tool.

MATERIALS AND METHODS

Study Design and Setting

This prospective observational study included 665 consecutive patients aged ≥18 years who underwent appendectomy with a preoperative diagnosis of AA at the Department of General Surgery.

Ethical Approval and Informed Consent

Ethical approval was obtained from the Institutional Review Board of Istanbul Medeniyet University, Goztepe Prof. Dr. Suleyman Yalcin City Hospital (Date: 26.12.2017, Decision no: 2017/0364). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants before enrollment.

Data Collection

For each patient, preoperative laboratory parameters, including WBC, CRP, NEU, NEU%, TBIL, and IBIL, were recorded. Radiological findings from CT and US, when performed as part of routine care, were documented. The predefined cutoff values used for laboratory markers in the diagnostic analyses are presented in Table 1.

Table 1.

Cut-off values of laboratory markers

Laboratory markers	Unit	Cut-off values
WBC	/µL	10,000
CRP	mg/dL	0.5
NEU	×109/L	6.5
%NEU	%	75.2
TBIL	mg/dL	1.2
IBIL	mg/dL	1.0

RPE: Resident physical examination; US: Ultrasonography; CT: Computed tomography; WBC: White blood cell count; CRP: C-reactive protein; NEU: Neutrophil count; %NEU: Neutrophil percentage; TBIL: Total bilirubin; IBIL: Indirect bilirubin; PPV: Positive predictive value; NPV: Negative predictive value.

Physical Examination Assessment

Patients whose laboratory and/or radiological results were not available at the time of clinical evaluation were independently examined by a specialist general surgeon and a general surgery resident participating in the study, without access to laboratory or imaging results. Physical examination findings were recorded separately and categorized as compatible, suspicious, or incompatible with AA based on standard clinical criteria.

Reference Standard

All appendectomy specimens were examined and reported by a single experienced pathologist to ensure consistency and minimize interobserver variability. Histopathological diagnosis was used as the reference standard for all diagnostic performance analyses.

Statistical Analysis

Categorical variables are presented as numbers and percentages, and continuous variables as mean±standard deviation (or median [interquartile range], where appropriate). The diagnostic performance of physical examination, laboratory markers, and imaging modalities was assessed by calculating sensitivity, specificity, PPV, NPV, and overall diagnostic accuracy, with 95% CIs. The reference standard was histopathological diagnosis. Comparisons of categorical variables, when required, were performed using the chi-square test or Fisher’s exact test. A two-sided p value <0.05 was considered statistically significant. Statistical analyses were performed using IBM SPSS Statistics (Version 26.0; IBM Corp., Armonk, NY, USA).

RESULTS

A total of 665 patients underwent appendectomy during the study period, including 242 women (36.4%) and 423 men (63.6%). The mean age of the study population was 34.9±14.2 years. The overall negative appendectomy rate was 5.8% (n=39), with a female-to-male ratio of 1:2 (13 women vs. 26 men) among these patients.

Based on intraoperative findings, the appendix was reported as normal in 14 patients (2.1%), AA in 306 patients (45.8%), phlegmonous AA in 254 patients (38.1%), and perforated AA in 91 patients (13.6%). In contrast, histopathological examination demonstrated a normal appendix in 39 patients (5.8%), AA in 76 patients (11.5%), phlegmonous or gangrenous AA in 417 patients (62.7%), and perforated AA in 46 patients (6.8%).

A neoplastic pathology was identified in 3 patients (0.45%), with neuroendocrine tumors detected in 2 specimens, while adenocarcinoma was encountered in 1 case. The patient with adenocarcinoma underwent completion right hemicolectomy in a second session, and curative surgical treatment was achieved.

Among the 39 patients with histopathologically confirmed normal appendices, intraoperative assessment favored AA in 29 patients (74.3%), whereas the surgeon reported the appendix as normal in 10 patients (25.7%).

A discrepancy between intraoperative and histopathological diagnosis of perforation was observed. While 91 patients (13.6%) were considered to have perforated AA intraoperatively, histopathological confirmation of perforation was present in only 27 of these patients (29.7%). Conversely, 19 patients (41.3%) with histopathologically confirmed perforation were evaluated as non-perforated during surgery and therefore did not receive postoperative antibiotic therapy at discharge. During follow-up, 17 patients (89.5%) completed 6-month follow-up, and 1 patient (5.3%) completed 1-month follow-up. One patient required rehospitalization due to abdominal pain and received antibiotherapy, and another patient developed serous umbilical discharge, which was managed conservatively with dressing. No complications were observed in the remaining patients. The distribution and discrepancy between intraoperative and histopathological diagnoses are illustrated in Figure 1.

Figure 1.

Intraoperative and histopathological diagnoses.

Preoperative imaging was performed in all patients. A CT scan was obtained in 482 patients (72.4%), US in 330 patients (49.6%), and both CT and US in 148 patients (22.2%). None of the patients underwent surgery without at least one imaging modality. Among patients with negative appendectomy, 26 patients (66.6%) received a CT scan, US was performed in 22 patients (56.4%), and 9 patients (23.1%) underwent both CT and US.

The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall diagnostic accuracy of PE, laboratory parameters, and imaging modalities are summarized in Table 2. The highest sensitivity values were observed for resident PE (RPE) at 95.2% (95% CI: 90.7–97.9%), followed by NEU at 87.9% (95% CI: 85.0–90.3%), CT at 85.7% (95% CI: 82.0–88.9%), and WBC at 85.6% (95% CI: 82.6–88.2%).

Table 2.

Diagnostic performance of clinical and laboratory tests

Test	Sensitivity (%)	95% CI sensitivity	Specificity (%)	95% CI specificity	PPV (%)	NPV (%)	Accuracy (%)
RPE	95.2	90.7–97.9	20.0	4.3–48.1	92.9	27.3	88.9
US	78.5	73.0–83.4	13.3	1.7–40.5	93.9	3.5	74.9
CT	85.7	82.0–88.9	36.4	17.2–59.3	96.2	11.9	83.3
WBC	85.6	82.6–88.2	30.8	17.0–47.6	95.2	11.8	82.4
CRP	69.0	65.1–72.6	33.3	19.0–50.2	94.3	6.3	66.9
NEU	87.9	85.0–90.3	33.3	19.0–50.2	95.5	14.6	84.7
%NEU	67.7	63.9–71.4	56.4	39.6–72.2	96.1	9.8	67.1
TBIL	22.9	19.7–26.4	89.7	75.8–97.1	97.3	6.8	26.9
IBIL	10.4	8.1–13.0	94.7	82.3–99.4	97.0	6.1	15.2

RPE: Resident physical examination; US: Ultrasonography; CT: Computed tomography; WBC: White blood cell count; CRP: C-reactive protein; NEU: Neutrophil count; %NEU: Neutrophil percentage; TBIL: Total bilirubin; IBIL: Indirect bilirubin; PPV: Positive predictive value; NPV: Negative predictive value.

The highest specificity values were obtained with IBIL at 94.7% (95% CI: 82.3–99.4%) and TBIL at 89.7% (95% CI: 75.8–97.1%). Among inflammatory markers, NEU% demonstrated a moderate specificity of 56.4% (95% CI: 39.6–72.2%), while imaging modalities showed lower specificity values, particularly US at 13.3% (95% CI: 1.7–40.5%) and CT at 36.4% (95% CI: 17.2–59.3%).

Overall diagnostic accuracy was highest for RPE (88.9%), NEU (84.7%), CT (83.3%), and WBC (82.4%), whereas bilirubin-based parameters demonstrated high specificity but limited sensitivity, resulting in lower overall accuracy rates.

DISCUSSION

Despite the increasing reliance on laboratory and imaging modalities, PE remains a fundamental component of the diagnostic process in AA. Accumulating contemporary evidence indicates that clinician experience and frequency of patient evaluation are key determinants of PE performance, with reported sensitivity values ranging from 80% to over 90% when examinations are performed by well-trained clinicians in high-volume centers [6]. In accordance with this body of evidence, our study demonstrated that RPE achieved the highest sensitivity among all diagnostic tools (95.2%). This superiority over specialist PE (SPE) may be attributed to the intensive patient exposure and continuous bedside involvement of surgical residents in university and training hospitals. Furthermore, interim analysis performed at the 18^th month of the study revealed similarly high RPE sensitivity (92%), supporting the consistency of this observation over time. These results underscore that structured clinical training and repeated examinations enhance diagnostic reliability, even in the era of advanced imaging.

In adults, CT has widely been regarded as the most accurate imaging modality for the diagnosis of AA [3]. Large-scale meta-analyses and population-based cohort studies published after 2020 report CT sensitivity between 85% and 98%, with specificity values often exceeding 90%, particularly when interpreted by experienced abdominal radiologists with access to clinical information [16]. Nevertheless, emerging real-world data highlight that diagnostic performance may decline in emergency settings, especially during off-hours, when imaging is frequently interpreted without adequate clinical context [15]. In our cohort, CT sensitivity (85.7%) was lower than the upper range reported in the literature. We believe this discrepancy is largely attributable to the fact that CT scans were commonly reported by off-site radiologists who did not have access to detailed patient anamnesis or PE findings. This situation reflects routine clinical practice in many centers and highlights that CT accuracy is not solely dependent on technology but also on integration with clinical data.

Due to its lack of ionizing radiation, low cost, and widespread availability, US continues to be recommended as a first-line imaging modality for AA [6]. However, published data demonstrate considerable heterogeneity in diagnostic performance, with reported sensitivity values ranging from 50% to over 90%, largely depending on operator expertise and institutional experience [14, 16]. Operator dependency and limited availability of expert radiologists outside regular working hours remain major limitations [15]. In parallel with these observations, the sensitivity of US was relatively low in our study. During off-hours, examinations were primarily performed by radiology residents, which may explain the reduced diagnostic yield.

Laboratory parameters remain essential adjuncts in AA diagnosis. White blood cell count is the most frequently used marker and is known for its high sensitivity but limited specificity [3]. Recent systematic reviews and pooled analyses report WBC sensitivity between 75% and 85%, with specificity generally below 60%, particularly at lower cutoff values [17]. Higher thresholds improve specificity but reduce sensitivity [11]. In our study, using a WBC cutoff of 10,000 cells/mm³, sensitivity was 85.6% and specificity 30.8%, findings that are consistent with contemporary literature. These results confirm that WBC alone is insufficient for definitive diagnosis but remains useful as part of a multimodal assessment, particularly in early disease stages when CRP may still be normal.

Instead of being a marker for early diagnosis, CRP has widely been recognized as an indicator of disease severity [11]. Evidence from recent prospective and retrospective cohorts suggests that CRP levels rise 8–12 hours after symptom onset and correlate more strongly with complicated or perforated AA than with uncomplicated disease [18]. Our results mirror these findings, demonstrating that CRP is not sufficient as an isolated diagnostic marker but serves as an important warning sign for delayed or complicated cases. Extremely elevated CRP levels in our cohort were associated with advanced inflammatory stages, reinforcing its role as a prognostic rather than diagnostic parameter.

As readily available inflammatory markers, NEU and NEU% have gained attention [7]. Several recent investigations indicate that NEU may provide higher sensitivity than WBC, while NEU% may offer improved specificity under certain cutoff values [19]. In our study, NEU demonstrated high sensitivity (87.9%) with low specificity (33.6%), whereas NEU% showed moderate sensitivity (67.7%) and specificity (56.4%). These findings support the view that neutrophil-based parameters are valuable supportive tools but insufficient for standalone diagnosis. Their greatest benefit lies in reducing diagnostic uncertainty when combined with clinical evaluation and imaging.

Hyperbilirubinemia has been increasingly investigated as a marker of complicated AA [20]. Pathophysiological and clinical studies published in recent years suggest that elevated TBIL and IBIL levels are associated with perforation and severe inflammation, likely due to endotoxemia and hepatic dysfunction secondary to bacterial translocation [21]. Consistent with these mechanisms, TBIL and IBIL exhibited the highest specificity and PPV among laboratory parameters in our cohort. Although sensitivity was limited, elevated bilirubin levels in our patients were strongly suggestive of advanced disease, highlighting their utility as warning markers for complicated AA.

Diagnostic scoring systems remain widely used, particularly in pediatric populations [22]. However, contemporary adult-focused studies increasingly emphasize that scoring systems should complement, rather than replace, clinical judgment [23]. The revised Jerusalem Appendicitis Guidelines emphasize individualized assessment based on symptoms, physical findings, and selective imaging rather than rigid score-based decision-making [6]. Our data lend further support to this paradigm, as the high sensitivity of RPE and the variable performance of laboratory and imaging modalities indicate that overreliance on any single diagnostic tool may be misleading. Clinical judgment remains central to accurate diagnosis.

Contemporary evidence suggests that surgeons tend to overestimate inflammatory severity intraoperatively, while early mucosal or intraluminal appendicitis may appear macroscopically normal [24]. In keeping with these observations, our study demonstrated significant discordance between intraoperative and histopathological findings, particularly regarding perforation. These findings highlight the limitations of macroscopic assessment and underscore the importance of histopathological confirmation and cautious postoperative management decisions.

This study has several limitations. First, although the prospective design and large cohort enhance the robustness of the findings, the study was conducted at a single tertiary referral center, which may limit external generalizability. Second, RPEs were performed by members of the study team, which may have introduced observer-related bias and potentially contributed to the high diagnostic sensitivity observed for resident assessment. However, this approach ensured standardized evaluation and reflects real-world practice in teaching hospitals where residents are primary frontline assessors. Third, radiological examinations were interpreted under routine clinical conditions by radiologists with varying levels of experience and without a standardized reporting protocol, which may have influenced the diagnostic performance of US and CT. Finally, laboratory cutoff values were derived from institutional standards rather than externally validated thresholds, and long-term postoperative outcomes were not systematically evaluated.

Conclusion

In conclusion, this prospective observational study demonstrates that PE remains a highly effective diagnostic component in AA diagnosis when performed in a structured and systematic manner. Among all evaluated diagnostic tools, RPE conducted by the study team showed the highest sensitivity, underscoring the diagnostic value of frequent bedside assessment and clinical engagement in high-volume emergency settings. Laboratory markers and imaging modalities provided important complementary information but exhibited variable accuracy when used independently. These findings support an integrated diagnostic strategy in which PE, laboratory parameters, and selective imaging are applied synergistically rather than in isolation. Even in the era of advanced diagnostic technologies, maintaining strong clinical examination skills remains essential for accurate diagnosis and optimal management of AA.

Ethics Committee Approval

This study was approved by the Istanbul Medeniyet University, Goztepe Prof. Dr. Suleyman Yalcin City Hospital (Date: 26.12.2017, Decision no: 2017/0364).

Informed Consent

Written informed consent was obtained.

Conflict of Interest

No conflict of interest was declared by the authors.

Financial Disclosure

The author declared that this study has received no financial support.

Use of AI for Writing Assistance

Not declared.

Authorship Contributions

Concept – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Design – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Supervision – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Resource – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Materials – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Data collection and/or processing – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Analysis and/or interpretation: ÖE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Literature review – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Writing – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA; Critical review – OE, FK, MAA, IML, CIE, ANTY, TE, HG, OA.

Peer-review

Externally peer-reviewed.