|Year : 1997 | Volume
| Issue : 11 | Page : 409-416
Accuracy of palpation and percussion manoeuvres in the diagnosis of splenomegaly
DS Chongtham1, MM Singh2, SP Kalantri3, S Pathak4
1 Medicine Dept, MG Inst of Med Scl, Sewagram, Wardha. Presently Senior Resident, Medicine Dept, PGIMER Chandigarh 160 012, India
2 Dept. of Community Med, PGIMER, Chandigarh 160 012., India
3 Medicine Dept., MG Inst of Med Scl, Sewagram, Wardha 442 102, India
4 Dept of Radiodiagnosis, MG Inst of Med Scl, Sewagram, Wardha 442 102., India
D S Chongtham
Medicine Dept, MG Inst of Med Scl, Sewagram, Wardha. Presently Senior Resident, Medicine Dept, PGIMER Chandigarh 160 012
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Chongtham D S, Singh M M, Kalantri S P, Pathak S. Accuracy of palpation and percussion manoeuvres in the diagnosis of splenomegaly. Indian J Med Sci 1997;51:409-16
|How to cite this URL:|
Chongtham D S, Singh M M, Kalantri S P, Pathak S. Accuracy of palpation and percussion manoeuvres in the diagnosis of splenomegaly. Indian J Med Sci [serial online] 1997 [cited 2015 Jul 1];51:409-16. Available from: http://www.indianjmedsci.org/text.asp?1997/51/11/409/11504
There have been rapid strides in the development of medical technology for the early diagnosis of diseases. However, the role of clinical examination still remains important in evaluation of patient. Indeed clinical evaluation is more powerful than laboratory investigations in arriving at diagnosis, prognosis and in planning treatment strategies for most patients.  Splenomegaly is a useful cilinical sign; few workers have reported the role of palpation and percussion manoeuvres in the bedside assessment of splenomegaly with varying reliability. ,,,,,,, The present study was conducted to see the clinical accuracy of some of these manoeuvres in diagnosing splenomegaly.
| ¤ Material and Methods|| |
This study was conducted in Kasturba hospital, Wardha between July 1993 and April 1994. Patients admitted in the Medicine ward due to hepatic, renal, cardiac and infectious problems were randomly chosen for the study. One of the investigators (DSC) blind to the details of history, general examination and laboratory investigation findings examined the study subjects for splenomegaly. Patients with left-sided pleural effusion, previous history of ascites or splenomegaly and known cases of splenomegaly were excluded from the study.
All the patients were examined in a fasting state or at least two hours after food intake by the following methods.
Supine Palpation : The patient lay down in supine position. The tips of the fingers of examiner's right hand were pressed gently just beneath the left costal margin and the patient was asked to take a long and deep breath as the palpation of a descending spleen was sought. If the spleen was not felt, the procedure was repeated by lowering the right hand 2 cms at a time towards the umbilicus until the examiner was confident that a massive spleen was not missed.
The other palpatory method elicited was Middleton's manoeuvre.  Both palpatory methods were scored on a five point scale of Hackett's semiquantitative system.  Three percussion manoeuvres elicited for diagnosing splenomegaly were Traube's space percussion,  Castell's manoeuvre  and Nixon's manoeuvre.  Traube's space percussion and Casteil's manoeuvre were noted on a five point scale: 1-definitely tympanitic; 2-probably tympanitic; 3-uncertain; 4-probably dull; 5-definitely dull. Grade three and above were taken for considering traube's space percussion as positive. Castell's manoeuvre was also considered as positive if on end expiration the percussion grade was four or more or if on inspiration percussion grade progressed by one or more grade towards definitely dull in those who had a percussion score of one, two or three on end expiration.
Abdominal ultrasound examination was performed on all the study subjects by a sonographer within twenty four hours of cilinical examination before starting treatment. He was blind to the patient's history and clinical examination results including palpatory and percussion manoeuvres. Spleen was diagnosed as enlarged if the sonographic cephalocaudal diameter was equal to 13 cms. or more. Quetelet's index (01) was calculated for each patient using the formula: QI Weight in Kg/Height in meter . 
Statistical analysis was done using Student's `t' test for comparison of demographic, morphometric and clinical variables among patients with or without splenomegaly. Sensitivity (true positive/(true positive + false positive)), specificity (true negative/(true negative + false positive) , positive predictive value (true positive/(true positive + false positive)) and negative predictive value (true negative/ (true negative + false negative)) were calculated for various manoeuvres taking the sonographic findings as gold standard. Receiver operating characteristic curves (ROC) were generated for various manoeuvres to compare their ability for detection of splenomegaly. For this the sensitivity and specificity of each palpation and percussion manoeuvres were evaluated for various score thresholds like for a cut off value of one, the definitely not palpable findings were considered as test negative and all others were considered as test positive, for a cut off value of two, definitely and probably not palpable findings were classified as negative and all others were taken as positive and so on. The upper left hand corner of the curve denoted a perfect diagnostic test, the 45 degree line represented a non-discriminate test (area=0.5) where the true positive rate equaled the false positive rate.
| ¤ Observations|| |
The study participants included 43 male and 37 female patients. Their mean age was 31.5 years (29.8 years males and 33.3 yaers for females) [Table 1]. None of the patients was obese (Quetelet's index less than 25 Kg./m 2 ). The mean weight of the study participants was 45.218.2 Kg and mean quetelet's index was 17.812.6 Kg/m 2 . Mean splenic size among the cases was 14.7 cms. And that among normal sized spleen group was 9.9 cms.
Accuracy of various bedside manoeuvres are shown in [Table 2]. The sensitivity of Middleton's and Castell's manoeuvres were similar (85.7%) and higher than other manoeuvres. It was least for Nixon's manoeuvre (66.7%). The specificity was highest (92.1%) with supine palpation and least (31.6%) with Castell's manoeuvre. The overall positive predictive value was highest (91.7%) with supine palpation whereas Middleton's manoeuvre showed hihest negative predictive value (84.6%).
Nixon's percussion manoeure was correlated with splenic size as shown in [Table 3]. The Quetelet's index was comparable among both Nixon's manoeuvre positive and negative groups. A significant difference in percussion distance (p<0.05) and splenic size (p<0.01) were also observed between the two groups.
The Receiver Operating Theistics curves of various manoeuvres are shown in [Figure 1]. With supine palpation a progressive decline in sensitivity from 98% to 50% was observed as the palpation threshold progressed from one to four while specificity increased from 58% to 95% in similar groups. The area under the ROC curve was observed to be 0.92 which is significantly greater than 0.5 of a non-discriminating test curve (p<0.05). The ROC curve for Middleton's manoeuvre was also generated based on the sensitivity and false positivity rates at various thresholds of palpation. The area under the curve was 0.93 which is comparable to that of supine palpation. The area under ROC curve generated from various percussion thresholds of Traube's Characterspace percussion was 0.74 which is greater than 0.5 for a nondiscriminating test indicating a better discriminating ability than chance alone. since obesity could influence the percussion note, stratification of the results was done according to Quetelet's index. However, all the patients were lean and thin and only 6 had Quetelet's index value between 20 and 25 Kg/M 2.Hence no statistically significant trend for Traube's space percussion was observed in relation to body habitus.
The area under Castell's manoeuvre using four percussion thresholds was 0.75. This manoeuvre resulted in a high proportion of false positive (69%) results. Castell's manoeuvre and Traube's space percussion showed comparable results in terms of discriminating splenomegaly from normal sized spleens.
The mean percussion dullness among those classified as positive Nixon's manoeuvre (9.7±1.5cms.SD) showed significantly larger spleen size (14.2±2.6 cms: SD) than those classified as negative Nixon's manoeuvre [Table 3].
| ¤ Disscussion|| |
It is generally agreed upon that palpable spleen is pathological. But there is no consensus regarding the cut off point between normal and enlarged spleen. Radio isotopic scintiscan and ultrasonographic recordings also differ in the standards of splenic size. Ultrasonography (USG) has most frequently been used for rapid and non-invasive assessment of splenomegaly. Spleen is said to be enlarged if its cephalo-caudal diameter exceeds 13 cms. 
Out of the various methods available for detection of splenomegaly, only three percussion manoeuvres (Nixon's, Castell's and Traube's space percussion) and three palpatory methods (Two-handed palpation with patient in right lateral decubitus, one handed palpation with patient in supine position and Middleton's manoeuvre) have been validated against USG or Scientigraphy. ,,,,,, Two retrospective studies revealed low sensitivity of palpatory methods ranging from 20%  to 28%  and specificity from 98% 
to 100%.  The correlation was based on autopsy and Scintigraphy was used as gold standard. This does not reflect the true sensitivity and specificity since the quality of clinical examination could not be judged from the records. However, five prospective studies [Table 4] using Scintigraphy , or operatives or USG , the sensitivity was ranging from 56% to 82% and specificity from 46% to 100%. The higher results could be because of the cautious nature of the clinicians in diagnosing splenomegaly since they were kept under scrutiny for the study. The present study was conducted by a clinician blind to the previous history of the study of the patients. There was no appreciable differences in relation to age, sex, weight, height and Quetelet's index characteristics between the patients with normal sized spleens and those with enlarged spleens.
Present study showed that palpatory methods did not differ much in their ability to detect splenomegaly. The sensitivity was higher (range, 78.6% to 85.7%) than reports from other previous prospective analysis of palpatory manoeuvres varying from 56% to 71 %. ,.. In 1967 Castell showed that percussion could reliably discriminate between a normal sized spleen from an enlarged one using scintiscan as gold standard.  The present study showed that Castell's manoeuvre had a sensitivity of 85.7% but generated high false positivity and hence less specific (31.7%). Nixon's method has not been scientifically validated except for one report  which showed a sensitivity of 59% and 94% specificity. The present study also showed a comparable sensitivity (66.7%) and specificity (81.6%). Traube's space percussion has long been regarded as unreSliable until 1989 when Barkun et als showed that this method had a sensitivity of 62% and specificity of 72%. He also noted that enlarged spleen is likely to be missed among obese patients and more false positive results were obtained if the patients were examined upto two hours after food. After excluding obese patients the sensitivity and specificity rate rose upto 78% and 82% respectively. In the present study the confounding effects of obesity and after food was taken care of since none of the patients was obese and examination was performed at least two hours after recent food intake.
A previous study  had shown that palpation was significantly a better discriminator among patients in whom percussion was positive. The study also inferred that if percussion dullness was absent, there was no need to palpate as palpation was a less reliable test. On the contrary, the present study showed that palpation was significantly better than percussion and the latter offered no extra advantage in bedside detection of splenomegaly.
The accuracy of various palpation and perccssion manoeuvres was also assessed by ROC curve analysis. this analysis also showed that supine palpation and Middleton's manoeuvre were similar in their accuracy for detection of splenomegaly. Similar observations for Castell's and Traube's space percussion were also seen. However, palpatory methods clearly showed a better discriminator over percussion for detection of splenomegaly.
The present study highlights the importance of physical examination in detecting splenomegaly. A careful examination performed at least two hours after food using palpatory techniques will definitely enhance the clinical acumen in detection of enlarged spleen which otherwise would have been routinely missed among nonobese individuals.
| ¤ Summary|| |
A study was conducted on 80 patients admitted in a teaching hospital to see the accuracy of two palpatory methods (Supine palpation and Middleton's manoeuvre) and three percussion methods (Traube's space percussion, Castell's and Nixon's manoeuvres) in the diagnosis of splenomegaly. Ultrasonographic findings were considered as gold standard for diagnosing splenomegaly. Mean age of study subjects was 31.5 years and mean Quetelet's index was 17.8±2.6 kg/m 2 . Sensitivity of Middleton's and Castell's manoeuvres was similar (85.7%) and higher than other manoeuvres. Nixon's manoeuvre had the least sensitivity (66.7%). Specificity was highest (92.1%) with supine palpation and least (31.6%) with Castell's manoeuvre. Supine palpation showed highest positive predictive value (91.7%). Receiver Operating Characteristics curves showed greater area with middleton's manoeuvre (0.93) followed by supine palpation (0.92), Castell's manoeuvre (0.75) and Traube's space percussion (0.74). the findings of the study suggests that palpatory methods like Middeton's manoeuvre and Supine palpation should be routinely used for diagnosing splenomegaly among non-obese individuals.
| ¤ References|| |
|1.||Crombie MD. Diagnostic process. J Coll gen Practit 1963;6:578. |
|2.||Riemenschneider PA, Whalen JR The relative accuracy of estimation of enlargement of the liver and spleen by radiologic and clinical methods. Am J Roentgenol Radiat Ther Nucl Med 1965;94:462-8. |
|3.||Halpren S, Coal M, Ashburn W et al. Correlation of liver and spleen size. Determinations by nuclear medicine studies and physical examination. Arch Intern Mad 1974;134:123-4. |
|4.||Westin J, Lanner LO, Larsson A, Weinfeld A. Spleen size in polycythaemia. A clinical and scintigraphic study. Acta Med Scand 1972;191:263-71. |
|5.||Ingeberg S, Stockel M, Sorensen PJ. Prediction of spleen size by routine radioisotope scintigraphy. Acta Haernatcl 1933;69:243-8. |
|6.||Zhang B, Lewis SM. Use of radionuclide scanning to estimate size of spleen in vivo. J Clin Pathol 1987;40:508-11. [PUBMED] [FULLTEXT] |
|7.||Sullivan S, Williams R. Reliability of clinical techniques for detecting splenic enlargement. BMJ 1976;2:1043-44. [PUBMED] [FULLTEXT] |
|8.||Barkun AN, Camus M, Green L et al. The bed-side assessment of spleflkk enlargement. Am J Med 1991;91:512-18. |
|9.||Barkun AN, Camus M, Meagher T et al. Splenic enlargement and Tiaube's space: how useful is percussion? Am J Med 1989;87:56266. |
|10.||Shaw MT, Dvorak V. Palpation of slightly enlarged spleens. Lancet 1973;1:317. [PUBMED] [FULLTEXT] |
|11.||Castelt DO. The spleen percussion sign: a useful diagnostic technique. Ann Intern Med 196767:1265-7. |
|12.||Nixon RK Jr. The detection of splenomegaly by percussion. N Engi J Med 1954;250:166-7. |
|13.||Neiderau C, Sonnenberg A, Muller JE, Erckenbrecht Jt, Scholten T, Fritsch WP. Sonographic measurements of the normal liver, spleen, pancreas and portal vein. Radiology 1983;149:537-40. |
|14.||Holzbach RT, Clark RE, Shipley RA, Kent WB, Lindsay GE. Evaluation of splenic size by radioactive scanning. J Lab Clin Med 1962;60:902-13. |
|15.||Colin 0, Christopher CE. Chamberlain's symptoms and signs in clinical medicine. 11th Edn. Oxford; ELBS with ButterworthHeinemann, 1987;102-3. |
[Table 1], [Table 2], [Table 3], [Table 4]