SCIENTIFIC ARTICLES Clinical and ultrasound correlation in the diagnosis of acute appendicitis: A retrospective study

Ultrasonography (USG) has been increasingly used to help in the diagnosis of acute appendicitis with the aim of minimizing both underdiagnosis and overdiagnosis of acute appendicitis. The aim of this study is to ascertain whether USG could change the decision making in the management of acute appendicitis. Methods This is a retrospective analysis of prospective data on 100 patients who underwent appendicectomy either open or laparoscopic. These 100 patients were divided in to two groups; those with equivocal appendicitis and those with strongly suspected appendicitis based on clinical assessment. Modified Alvarado Score (MAS) was used for clinical assessment.


Introduction
Ultrasonography (USG) has been increasingly used to help in the diagnosis of acute appendicitis with the aim of minimizing both underdiagnosis and overdiagnosis of acute appendicitis. The aim of this study is to ascertain whether USG could change the decision making in the management of acute appendicitis.

Methods
This is a retrospective analysis of prospective data on 100 patients who underwent appendicectomy either open or laparoscopic. These 100 patients were divided in to two groups; those with equivocal appendicitis and those with strongly suspected appendicitis based on clinical assessment. Modified Alvarado Score (MAS) was used for clinical assessment.
All of them had USG as the sole mode of imaging. The accuracy of each mode of assessment was confirmed by macroscopic and/or microscopic appearance of the appendix. Based on the findings it was judged whether performing USG has made any changes to the clinical decision.

Results
There were 61 males and 39 females with a mean age of 25 (7-86) years. Total number of patients admitted with modified Alvarado score (MAS) of 2-6 was 39 and those with MAS of 7-9 was 61. True positive cases of acute appendicitis in the former group were 33. While ultrasonography detected acute appendicitis correctly in 30 patients, only 3 patients were correctly diagnosed clinically. Out of 59 true positive cases of acute appendicitis in the latter group, USG detected only 41 cases whereas all 59 patients were correctly identified clinically. While USG has helped change the clinical decision in the equivocal appendicitis, clinical decision takes precedence over USG in patients with typical appendicitis.

Introduction
Acute appendicitis is one of the commonest acute surgical emergencies and the diagnosis can often be made clinically (1,2). Although in its typical presentation, diagnosis of acute appendicitis poses little challenge, it is not always straightforward in atypical presentation and thus patients may end up in either negative appendicectomies (i.e. false positive cases) or the surgery may be delayed with adverse outcome (2). Alvarado introduced a scoring system to help minimize either eventuality which was later modified by Kalan et al (3,4). Adding imaging to clinical diagnosis has been considered to improve the positive predictive value of acute appendicitis thus minimizing the negative appendicectomy rates (5,6). Both Ultrasound scan and computed tomography have been shown to be effective in this regard.
The aim of this study is to find out accuracy of clinical and radiological assessment in diagnosing acute appendicitis and thus to ascertain whether performing preoperative ultrasonography would change the clinical decision in the management of suspected acute appendicitis. Since the body habitus is considered to have effect on the accuracy of the abdominal ultrasound scan findings, body mass index of each patient too was calculated.
All 100 patients had appendicectomy performed irrespective of finding a macroscopically normal looking appendix intraoperatively. This approach had to be undertaken to avoid future diagnostic confusion by leaving a macroscopically normal appendix in case of open appendicectomy. The same policy was followed in our unit in laparoscopic approach too in the absence of alternative diagnosis since microscopic appendicitis in the background of macroscopically normal appendix has been a recognized entity although its clinical significance and natural history is not well defined (7). We, however, were equally aware of both sides of the debate in removing a macroscopically normal appendix (8,9).

Method
This was a retrospective analysis of a prospectively maintained data and hence no ethical clearance was sought for. Records  Ultrasound was performed by either a consultant radiologist or a medical officer of radiology unit who has adequately been trained for this purpose. Ultrasound detection of a non-compressible blind ending tubular structure of more than 6mm in diameter noted in the right iliac fossa (RIF) or right hypochondrium (RHQ), free fluid in RIF, presence of appendicolith and or increase appendicular vascularity were considered positive for acute appendicitis radiologically. Failure to spot appendix or detection of normal appendix was regarded as criteria for negative ultrasound scan.
These 100 patients included 3 categories of patients. First group were those who scored 7 or more points in MAS and they were considered as having clinical appendicitis and proceeded to appendicectomy irrespective of ultrasound scan findings whereas others with a lesser score had appendicectomy done only if the ultrasound scan made a positive diagnosis. The third group comprised those who had appendicectomy despite MAS of 4-6 and negative ultrasound scan findings. This category of patients had been observed without starting antibiotics for progression of symptoms. Since their score either remained the same or increased over the ensuing 1-2 days they too were considered clinically positive for acute appendicitis.
Those who proceeded to surgery had their BMI calculated and had the surgery done either laparoscopically or by conventional open method depending on the feasibility of the emergency list. All laparoscopic appendicectomies were undertaken by a consultant surgeon while open appendicectomy was performed either by a consultant surgeon or by an experienced SHO grade doctor. Definitive diagnosis of appendicitis was made either macroscopically intraoperatively and/or histologically in case of less obvious macroscopic findings. Macroscopic appearance of a non compressible distended appendix with red serosa and prominently distended vessels coursing along appendix constituted grossly inflammed appendix whereas only distended appendix with dilated vessels coursing along the appendix was considered minimally inflammed appendix.
Transmural infiltration of appendix with neutrophils or scattered neutrophils in appendeceal mucosa was considered histological proof of appendicitis.

Results
Over the study period, 100 patients aged 7 years to 86 years had preoperative ultrasonography. Overall, 75 patients of 100 were reported positive by ultrasonography while only 71 of them had appendicitis. Therefore four patients had false positive findings by sonography. Out of 25 negative sonography, 21 patients had inflammed appendices while only 4 had normal appendices. Therefore ultrasonography in our study had an overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of 77%, 50%, 94.6%, 16% and 75% respectively ( Table 2).
Clinical evaluation was positive for appendicitis in 66 patients while inflammed appendices were discovered in 62 patients. Thirty out of 34 clinically negative patients had inflammed appendices. Overall, our results show sensitivity, specificity, PPV, NPV and accuracy of 67.4%, 50%, 94%, 11.8% and 66% respectively for clinical examination (Table 2).
While there were 39 patients (39%) who belonged to the normal weight group, the number of underweight and overweight patients were 48 (48%) and 12 (12%) respectively. Only one patient belonged to class 1 obesity group, whose USS was inconclusive (hence taken as negative) due to excess body fat.

Discussion
Decision to undertake appendicectomy is based on clinical grounds and appendicectomy is therefore undertaken without preoperative imaging in most cases. However, this approach is likely to result in negative appendicectomies especially in atypical presentation. It is widely believed that preoperative imaging with USG  and or computerized tomography (CT) scan help reduce negative appendicectomy rate (5,6). We did retrospective analysis to find whether performance of routine USG would help change clinical decision to undertake appendicectomy.
Hundred patients included in our study were analyzed in terms of two groups namely those belonging to MAS of 2-6 where suspicious for acute appendicitis is low and MAS of 7-9 in which probability of acute appendicitis is high. Therefore it could be assumed that those 34 patients had the operation undertaken timely due to the additional evidence provided by the USG. Accordingly, USG in this category of patients has allowed not only timely intervention with overall presumed benefit of reduced hospital stay it also has allowed to minimize potential complications associated with delayed diagnosis. Accordingly, it could be argued that ultrasonography, in our study, has positively contributed to change the management of 30 patients (30%) in otherwise clinically unsuspected acute appendicitis. Although this approach, ie, USG directed surgical decision in this category has given rise to negative appendicectomy rate of 11.8 % (4/34) it still lies within the acceptable negative appendicectomy range (10). Moreover it is far below the negative appendicectomy rate of 40% (2/5) observed for clinically guided appendicectomy in this cohort of patients.
Ultarasonography, in evaluating this group of patients had shown superior sensitivity, PPV, NPV and accuracy compared to those of clinical examination (Table 3).
By contrast, USG paints a different picture in assessing patients having high probability of acute appendicitis. Total number of patients belonging to this category (MAS of 7-9) was 61 (61%) and all of them underwent appendicectomy on clinical diagnosis alone. All, except 2 patients, had inflammed appendices.
Ultrasonography, however, picked up inflammed appendix in only 67% cases (41/61), all of whom had appendicitis. Only 2 of remaining 20 patients had normal appendix while all others (18 patients) had clearly inflammed appendices. The assessment of patients with USG in this cohort of patients seems to be inferior to clinical assessment as shown by high sensitivity, PPV and NPV for clinical examination, each rated above 95% (Table 4).
Although USG has marginally higher PPV and over clinical examination, it has got very low NPV of 10%. Therefore normal USG in this cohort of patients is unreliable and as such negative findings on USG unlikely to change the management in this category (with AS of 7-9) of patients. This finding in our study supports the notion of performing appendicectomy on clinical grounds alone irrespective of USG findings in those admitted with typical history of acute appendicitis. While, overall sensitivity and accuracy of sonographic evaluation is better than clinical judgment, both mode of assessments have almost similar specificity, PPV, and NPV in the evaluation of patients ( Table 2).
The sensitivity and specificity of our study are comparable to those of Ozkan et al (11) who had sensitivity and specificity of 71.2% and 46.7% respectively.
Others (12,13) have reported sensitivity, PPV, and accuracy similar to our study; however, their specificity and NPV stand at 78% and 33-39% respectively. Literature survey has shown a range of sensitivity and specificity for USG indicating the gross operator dependency of the final result of USG. In our study USG was undertaken by 2 doctors with different level of experience whose accuracy has been in the range of 68-85%.
Furthermore amount of abdominal fat is known to interfere with the accuracy of the USG. However, majority (87%) in our study was in the underweight or normal weight group, while 12% were in the overweight range. None of them posed any problem in the sonographic assessment of the abdomen. Out of 12 overweight patients only 3 patients had discordant ultrasound scan and macroscopic/microscopic appearance, a finding applicable to patients in normal body weight group too. Only one patient (1%) had BMI of 32.2Kg/m2, interfering with the accuracy where USG was reported inconclusive due to excessive abdominal fat. Therefore it could be reasonably assumed that physical nature of patients enrolled in the study had little impact on the final result of our study.
However, the ultrasound scan machine used does not belong to the current generation and hence the sonographic assessment is likely to be suboptimal. This perhaps had some bearing on the final result considering the significantly low overall NPV (16%) of our study, whereas most of other studies reported NPV ranging from 31%-97% (11,12,13,14).
Overall negative appendicectomy rate in our study was 16% which is comparable to others (13) although a rate as low as 3%-6% has been reported in some studies (15,16).

Conclusion
Ultrasound scan is an important imaging modality in evaluating patients with suspected acute appendicitis with equivocal presentation with MAS of less than 6 points. It is likely to deliver important and decision making findings to the overall management of patients in this category. Therefore performing USS in these patients is advocated.
Patients getting admitted with typical history of appendicitis with MAS of 7-9 are unlikely to benefit from the findings of USS and performing scan in this group is perhaps not necessary.

Limitations
This study takes into account only the patients who actually had appendicectomies, however, in reality there were some patients with right iliac fossa pain whose clinical assessment and USG were both negative for acute appendicitis (i.e., seemingly true negative cases) and hence discharged without being enrolled into the study. Although this has undoubtedly contributed to relatively low overall specificity, NPP and accuracy of USG and clinical assessment, we believe that overall conclusion of the study is unlikely to be affected by this exclusion.
Moreover, in this study, failure to spot an appendix ultrasonically constituted a negative ultrasound scan. Although failure in identification of the appendix ultrasonically does not necessarily mean normal appendix, this could, at least, be argued as inability to provide an ultrasonic diagnosis of acute appendicitis for the purpose of this study. Therefore, this non conclusive result by USG is likely to affect negative predictive value to certain extent.