Objective: To determine the diagnostic accuracy of a protocol based on modified Alvarado score and ultrasonography in equivocal cases in appendicitis at Kenyatta National Hospital.
Design: A prospective observational study
Setting: KNH accident and emergency and general surgery departments
Patients and methods: Patients presenting with suspected appendicitis will be scored using the modified Alvarado score. Patients who score 7 and above will proceed to surgery while those who score the equivocal range of 4 to 6 will undergo ultrasound scanning for suspected appendicitis as described by Puylaert. Confirmation of appendicitis will be based on histology.
Main outcome measures: the main outcome measure will be the negative appendicectomy rate.
Data analysis: The sensitivity, specificity, positive predictive value and negative predictive value of the modified Alvarado score, Ultrasonography in equivocal cases and their combined use will be calculated.
Appendicitis was first described in 1886 by Reginald Fitz. It remains the most common cause of acute abdomen requiring surgical intervention both in Kenya and elsewhere.1,2,3,4,5,6 Accurate diagnosis of acute appendicitis remains a major challenge the world over and is perhaps more pronounced in constrained resource setups.7,8 Accuracy in diagnosis of acute appendicitis by clinical acumen has been found to be largely dependent on experience.7 Aids to assist in diagnosis of acute appendicitis exist but many are complex, expensive and unavailable especially in poor settings. Numerous scoring systems have been devised to aid the clinician.9 Perhaps the best well known is the modified Alvarado score. Various imaging modalities are available but their use is largely dependent on levels of resource. Ultrasonography has been used and studied widely in diagnosis of appendicitis.10 The aim of this study is to determine the diagnostic accuracy of a protocol based on the use of modified Alvarado score and ultrasonography in the equivocal cases in diagnosis of acute appendicitis with the main outcome measure being the negative appendicectomy rate.
Reginald Fitz first described the entity appendicitis in 1886. Diagnosis of acute appendicitis has remained a challenge despite great advances in technology. Jones concluded that a negative appendicectomy rate of 20% has been generally accepted in a review of trends in management over thirty years.11 It is known that negative appendicectomy rates vary widely principally due to differences in experience. John et al found the sensitivity of clinical acumen in diagnosis of acute appendicitis to range between 71% and 97% depending on the experience of the clinician .7 Various studies have demonstrated that clinical acumen remains the mainstay in the diagnosis of acute appendicitis. Appendicitis in the young offers unique challenges with higher rates of perforation being observed with decreasing age.8 A study by Macklin et al showed that clinical diagnosis was superior to the modified Alvarado score in children.12
Pruekprasert found that in the hands of an experienced surgeon clinical acumen was superior to either modified Alvarado or CRP measurements. CRP measurements and the Alvarado score were quoted to be of value to the inexperienced surgeon.13 Disparities in clinical acumen will continue to exist since the apprentice nature of surgical training is both time and training dependent.
Various scoring systems have been devised to assist in improving accuracy in diagnosis of appendicitis. They include Alvarado, Teicher, Christian, Fenyo and Lindberg. 9
The Alvarado score puts a score to the common symptoms, signs and laboratory typically found in appendicitis. It was first described by Alvarado in 1986. 14
The modified Alvarado scoring system was found by the Abdominal Pain Study Group to meet the set criteria in terms of reduction in morbidity and mortality in reevaluation of published data. 9
This scoring system gives points for symptoms (migration of pain, anorexia, and nausea), physical signs (right lower quadrant tenderness, rebound tenderness, and pyrexia), and laboratory values (leukocytosis). Whether to include a right to left shift is dependent on the laboratory in use. The modified Alvarado score does not include the shift.
Modified Alvarado score
Migration of pain
Elevated temperature (≥ 37.3ºC)
A prospective study of 116 patients by Ongaro at Kenyatta National Hospital in 2005 found that use of modified Alvarado score would have reduced negative appendicectomy rates from 25% to 11.2%. The sensitivity of the scoring system was found to have a sensitivity of 91 %. 15
The modified Alvarado score has been found a useful tool for admission criteria with one study giving a negative appendicectomy rate of 12.5%.In this study by Al Qahtani in Saudi Arabia, no patients with a score less than 4 had appendicitis.16
Kalan et al found that high Alvarado score was an easy and satisfactory aid in diagnosis of acute appendicitis in children and men. There was however an unacceptably high false positive among women of 33% versus 22% in the others. 17
Alvarado scoring in children is a useful tool in taking the decision for admission in suspected acute appendicitis. 18
Khan and Rehman found a negative appendicectomy rate of 15.6% and a positive predictive value of 84.3% in a study of 100 patients. They recommended Alvarado score as an easy, simple and complementary tool for the diagnosis of acute appendicitis especially for the junior surgeons.19
Ahmed et al reported that the score had a positive predictive value of 98.1% in their study of 100 patients with acute appendicitis. 20
McKay and Shepherd did a study on 150 patients. Alvarado score below 3 was found to have a sensitivity of 96.2%. The sensitivity and specificity of the score above 7 was 77% and 100% respectively. 21
A study by Denizbasi et al showed there was no statistical difference between the use of Alvarado score by the emergency medicine residents and the general surgery residents in terms of suspecting the diagnosis of appendicitis. Overall sensitivity was 95.4% and a specificity of 45.7%. 22
Ultrasonography in appendicitis
Ultrasonography as an imaging modality is widely available in the country. Anecdotal evidence suggests that it can be found in most district and provincial hospitals and also among private radiologists and radiographers in Kenya.
Pulyert described a graded compression technique for evaluating the appendix with transabdominal sonography in 1986.His study had 60 consecutive patients with suspected acute appendicitis. The inflamed appendix was visualized in 25(89%) of 28 patients with confirmed appendicitis. The appendix was not visualized in the 32 patients without appendicitis. Ultrasonography was also able to pick 6 out of 7 perforated appendices. Parameters to check for included an outer diameter of more than 6mm, aperistalsis, noncompressibility and periappendiceal fluid. 23
Ultrasonography offers the added advantages of non invasiveness, short acquisition time, lack of radiation exposure and the potential to diagnose other cause of acute abdomen especially in young women. Poortman et al suggested that ultrasonography should be incorporated as a first line imaging modality for the diagnosis of acute appendicitis in adults.24
A study by Barloon et al showed that graded compression ultrasonography had a sensitivity of 66% and a specificity of 95% in first and second trimester. 25
Ultrasonography has been described as a useful tool in diagnosis of appendicitis in the young. 26
Balthazar et al correlated CT and US in a 100 patients and concluded that the sensitivity of CT was higher (96% for CT, 76% for ultrasound), while the specificity was comparable (89% for CT, 91% for ultrasound), yielding a higher accuracy for CT (94% versus 83%).The specificity of US in this study was actually higher in this study. 27
Nicolas et al in their study population of 125 patients concluded that a threshold 6mm appendix under compression is the most accurate US finding for appendicitis with a high positive and negative predictive value. Their study evaluated the predictive values, sensitivity and specificity of US, Doppler US and laboratory findings in appendicitis. A finding of an appendix greater than 6 mm had a PPV, NPV, sensitivity and specificity of 98 %.28
Prospective studies have shown that the overall accuracy of US in diagnosing acute appendicitis ranges between 87-96% (sensitivity 75-90%, specificity 86-100%) with positive and negative predictive values of 91-94% and 89-97% respectively. 29,30,31,32 In experienced hands US significantly improves diagnostic accuracy in suspected appendicitis while reducing the negative laparatomy rate to 8-15%. 33
A prospective study by Blank and Braun found that as a result of sonographic detection or exclusion of other disease, ultrasonography facilitated the often difficult clinical differential diagnosis of right lower abdominal pain.33
Sim et al studied 80 patients with equivocal signs and symptoms of acute appendicitis.They reported a specificity of 100% and an overall sensitivity of 90 %.35
Poortman et al in their study population of 199 patients compared graded compression technique ultrasonography and unenhanced single-detector helical CT in diagnosis of acute appendicitis. They reported similar accuracy. Of note, the imaging was done by both imaging radiologists and general radiology staff.36
Gamanagatti et al showed that CT scanning had better accuracy although the specificity of ultrasonography was still 100% in this study.37
Ultrasonography was a more accurate diagnostic method than IL- 6 serum concentration, a laboratory marker with the highest diagnostic in a study by Groselnj-Grenc et al. They concluded that it should be a part of the diagnostic procedure for acute appendicitis in children. 38
The addition of various operator dependent ultrasonography techniques have been shown to further improve its utility with sensitivity and specificity approaching 100 %. 39
The results of the study by Caroline et al showed that the diagnostic performance of US does not differ from that of unenhanced multi–detector row CT in the detection of both acute appendicitis and alternative diseases, regardless of radiologist expertise in gastrointestinal imaging or patient sex, age, and body size. 40
Bendeck et al in a retrospective study of 462 patients reported a negative appendicectomy rate of 8% in women who underwent ultrasound scanning as compared to 28% in those who did not undergo any preoperative imaging. Ultrasonography had a sensitivity of 77% and positive predictive value above 92%.41
Brigand et al have suggested that imaging is useful if a scoring system for appendicitis places the probability as doubtful. 42
Preeyacha et al in their retrospective study concluded that ultrasonography has a negative predictive value of 95.1 %.43 As such one can infer that ultrasonography would be of great value in decision making on equivocal cases.
In patients with suspected acute appendicitis, US examination with the option of additional CT significantly lowers the negative appendectomy rate as compared with the clinical acumen alone, without adverse effects on the perforation rate or the in-hospital delay. The negative appendicectomy rate was 6% in this study. Combined ultrasonography and CT imaging for equivocal cases had sensitivity of 96.7% 44
The overall accuracy of ultrasonography was higher than that of surgeon’s clinical impression in a study by David et al. 45
Alvarado score and other modalities
Combination of the Alvarado scoring with other diagnostic modalities has been reported to show good results. The score was combined with laparascopy and gave favourable results of 0% negative appendicectomies compared to 18% in the control group. The algorithm was especially useful in the female population. 46
McKay and Shepherd in their study recommended the use of adjunctive CT scans in patients with equivocal Alvarado scores between 4 and 6. 47
The addition of ultrasonography in negative or equivocal cases using the Alvarado score decreased the false negative rate by 75% in a study by Stephens and Mazzucco.In their study, combination of both modalities reduced false positives to zero. 48
A randomized control trial of ultrasonography in diagnosis of acute appendicitis incorporating the Alvarado score found had a sensitivity and specificity of 94.7% and 88.9%. The decision to do ultrasonography was based on the Alvarado score. Patients in the intervention ultrasonography-Alvarado group had a shorter mean time to operation than the controls. 49
Debnath et al in their study concluded that graded compression ultrasonography is an accurate means of diagnosing or excluding acute appendicitis in clinically equivocal cases and it is of great value in establishing alternative diagnosis. 50
What is the overall negative appendicectomy rate at KNH following the use of a protocol based on modified Alvarado score and ultrasonography in equivocal cases in diagnosis?
Acute appendicitis remains the most common cause of acute abdomen requiring surgical intervention.1,5,6 Accurate diagnosis still largely depends on the experience of the clinician.8,25 Various scoring systems and imaging studies are currently in use to increase accuracy of diagnosis while avoiding increase in complication rates. A negative appendicectomy rate and complication rates are common outcomes used in surgical audits.
The negative appendicectomy rate in our setup is 25% overall and higher in women. 15
The modified Alvarado score is an easy and reproducible score whose utility has been found among clinicians at different levels of experience. 17, 18, 19, 46 Ultrasonography is a widely available modality in our setting and has been shown to be useful in many studies. Overall sensitivity and specificity was reported as 85% and 92% respectively in a metaanalysis by Orr et al. 46
There is paucity of data and studies on combining the modified Alvarado score and ultrasound scanning in management of acute appendicitis in our setup.
This prospective observational study seeks to establish whether the combined use of a simple clinical scoring system and ultrasonography in a protocol can lead to lower negative appendicectomy rates at KNH. Positive results will encourage an affordable protocol based criteria in diagnosis of acute appendicitis at KNH and across the country.
To determine the diagnostic accuracy of a protocol based on modified Alvarado score and ultrasonography in equivocal cases in diagnosis of appendicitis at KNH
1) To determine the sensitivity, specificity and predictive value of modified Alvarado scoring for acute appendicitis at KNH.
2) To determine the sensitivity, specificity and predictive value of ultrasonography for equivocal cases of appendicitis at KNH.
3) To determine the overall negative appendicectomy rate.
4) To determine the overall accuracy of a protocol based on modified Alvarado score and ultrasonography in equivocal cases for diagnosis of appendicitis at KNH
PATIENTS AND METHODS
The study will be conducted at KNH accident and emergency and surgical wards. KNH is a national teaching and referral hospital in Kenya. It serves Nairobi and its environs and also serves as the referral center for the country and its neighboring countries.
All patients above 5 years of age with suspected appendicitis seen at the accident and emergency department and in the surgical wards.
Prospective observational study
A previous study by Neford had a sample size of 116.
Appendicitis had a prevalence of 32.5% and 37.5% in the retrospective and prospective arms respectively in the study by Mungai.1 The average is taken 35%.
This is an observational study to evaluate diagnostic accuracy.
Buderer’s formula was used to calculate the required sample size for given values of specificity, sensitivity and absolute precision:
SN= anticipated sensitivity
Z1-a/2= Statistic for the level of confidence of 95%, 1.96
d = absolute precision desired (half the width of the confidence interval)
The prevalence estimates for appendicitis were taken to be 35%, an average for the retrospective and prospective arms of a study carried out in our settings. Taking the sensitivity and specificity of the score, the ultrasound and both combined to be above 70%, we calculated the required sample size for 0.1 value of precision
For sensitivity or specificity above 90%, sample sizes of about 100 or even less will achieve a considerable precision of 10%. Using the formulae above Malhotra and Indrayan(2010) have developed a nomogram where different samples sizes can easily be read of the scale for different values of prevalence, sensitivity and precision.
The subset of patients scoring 4 to 6 on the modified Alvarado score will be a minimum of 30 to allow for analysis.
All patients above 5 years with suspected acute appendicitis
1) Patients with generalized peritonitis
2) Patients with previous abdominal surgery
3) Patients with blunt or penetrating abdominal trauma
All patients who meet the eligibility criteria and consent to be recruited into the study will undergo diagnosis according to a protocol incorporating the modified Alvarado score and graded compression ultrasonography.
Consecutive patients above 5 years with suspected appendicitis will be recruited at the Kenyatta National Hospital accident and emergency department over a period of 6 months. All the patients will have a total blood count done. The results will be used together with the clinical history and examination to get the modified Alvarado score.
Patients who score 7 and above will undergo surgery for appendicitis. No further imaging will be done.
Patients who will score between 4 and 6 will undergo graded compression ultrasonography as described by Puylaert. The ultrasonography will be done by qualified sonographers and diagnostic radiology residents. The right iliac fossa findings will be recorded in a standard form that will contain the parameters described by Puylaert. The ultrasonography findings will be recorded as normal, appendicitis, indeterminate or other pathology. They will then proceed according to an algorithm. (See appendix I)
Patients with another abdominal pathology picked on ultrasound will be noted for subanalysis.
Patients with a score of 3 and below will be discharged or observed as per the clinician’s prerogative.
Confirmation of the diagnosis of appendicitis will be based on histology.
The ultimate decision on appendicectomy will rest on the admitting clinician. These patients will be included in the analysis.
DATA MANAGEMENT AND ANALYSIS
Data will be collected using a standard form for the biodata, history and examination findings, and the modified Alvarado score will be calculated. Ultrasonography findings will be entered into a standard form. The data will then be entered in customized Ms Access database with in-built checks to minimize on data entry error. Statistical analysis will be done using SPSS version 15.0 and Stata v10.
Descriptive statics will be calculated for the patient population. Taking the histological findings as gold standard, specificity and sensitivity and their 95% confidence intervals will be estimated for a) the score b) the ultrasound c) both score and ultrasound combined. Their positive predictive values and negative predictive values will also be computed. Negative appendectomy rates and the expected reduction achieved by combining the 2 diagnostic tools will be calculated.
Approval to carry out the study will be sought from the Surgery Department University of Nairobi and the Kenyatta National Hospital Ethics and Research Committee.
Patients recruited into the study will have given a signed consent after a clear explanation of the nature and purpose of the study. Patients who decline to consent will not have their treatment jeopardized in any way.
Minors will consent through their next of kin or guardian.
Ultrasonography will not be carried out by a designated person. This will however be mitigated by the use of standard ultrasonography descriptions using Puylaert’s graded compression method. It is felt that this may also increase the external validity of the study.
Proposal writing and submission for ethical approval
December 2009 to February 2011
Data collection and analysis
April 2011 to September 2011
Presentation of results
Research fees (KNH/ERC)
Stationery, printing and binding@100 per patient
Total blood counts@200 for 100 patients
Ultrasonography @1500 for 30 patients
The study will be funded by the principal investigator.
Cite This Work
To export a reference to this article please select a referencing stye below: