Tonsillectomy Rate Research Articles

Variation in paediatric tonsillectomy rates between Scottish health board areas, 2001-2018: is socio-economic deprivation to blame?

Tonsillectomy is one of the commonest operations in children. Routinely collected national data were used to assess variations in the paediatric tonsillectomy rate across Scotland, and to determine if socio-economic deprivation is the cause. The Scottish Morbidity Records were reviewed for all children (0-16 years) undergoing tonsillectomy from 2001 to 2018. The mean annual tonsillectomy rate was 2.64 per 1000 children. Rates in each health board area varied from 1.24 to 3.9 per 1000. Half of this variation resulted from transfers between regions. There was a 1.75-fold difference between tonsillectomy rates in the most and least deprived population quintiles, but this did not account for the geographical variation. Half the variance in paediatric tonsillectomy rates is associated with children being transferred between regions for treatment. After accounting for this, there is a 1.5-fold difference in rate between health board areas, which is not related to socio-economic deprivation and is currently unexplained.

A 4-year retrospective study of tonsillectomy rate and admission rate of tonsillitis and complications in the East of England and nationally.

Guidance set out by Scottish Intercollegiate Guidelines Network (SIGN) sought to restrict the frequency of elective tonsillectomy in reaction to the recognised and substantial risks of the procedure, namely pain and bleeding. With stricter criteria in place before a patient can undergo tonsillectomy, patients must endure more episodes of tonsillitis than was previously the case. The cost of managing tonsillitis and its complications as an inpatient is substantial to the NHS and also to the economy as a whole in missed work days. The authors sought to establish if the reduced rate of tonsillectomy performed had any effect on the rate of acute hospital admission for tonsillitis or the associated abscesses-peritonsillar, parapharyngeal and retropharyngeal. A retrospective multi-centre study reviewed admission data across a 4-year period. The rate of tonsillectomies performed for recurrent tonsillitis across four trusts in Mid-Essex was compared with the number of admissions for tonsillitis, peritonsillar, parapharyngeal and retropharyngeal abscesses. National hospital episode statistics data over a 5-year period was also analysed. For the regional data in 2015, 979 tonsillectomies were performed across the four centres, reducing to 875 in 2018. There was no trend in the rates of acute tonsillitis requiring admission but the rates of peritonsillar abscess increased from 156 to 192 cases per year in the same period. This correlation was found not to be statistically significant in the measured sample size. The National hospital episode statistics data showed a significant correlation between tonsillectomy rates and admissions from complications of tonsillitis as well as the associated abscesses. This study shows that the reduced tonsillectomy rate was correlated with an increased number of admissions with peritonsillar abscess regionally. Nationally reduced tonsillectomy rate is significantly associated with increased admissions with tonsillitis and all its complications. A decreased rate of tonsillectomy may be increasing the rate of serious tonsillitis. This has an impact on patient morbidity, an increasing financial burden on the NHS and the UK economy.

Differences in Tonsillectomy Use by Race/Ethnicity and Type of Health Insurance Before and After the 2011 Tonsillectomy Clinical Practice Guidelines

To evaluate whether differences in pediatric tonsillectomy use by race/ethnicity and type of insurance were impacted by the American Academy of Otolaryngology-Head and Neck Surgery's 2011 tonsillectomy clinical practice guidelines. We included children aged <15years from Florida or South Carolina who underwent tonsillectomy in 2004-2017. Annual tonsillectomy rates within groups defined by race/ethnicity and type of health insurance were calculated using US Census data, and interrupted time series analyses were used to compare the guidelines' impact on utilization across groups. The average annual tonsillectomy rate was greater among non-Hispanic white children (66 procedures per 10 000 children) than non-Hispanic black (38 procedures per 10 000 children) or Hispanic children (41 procedures per 10 000 children) (P < .001). From the year before to the year after the guidelines' release, tonsillectomy use decreased among non-Hispanic white children (-11.1 procedures per 10 000 children), but not among non-Hispanic black (-0.9 procedures per 10 000 children) or Hispanic children (+3.9 procedures per 10 000 children) (P < .05). Use was greater among publicly than privately insured children (75 vs 52 procedures per 10 000 children, P < .001). The guidelines were associated with a reversal of the upward trend in use seen in 2004-2010 among publicly insured children (-5.5 procedures per 10 000 children per year, P < .001). Tonsillectomy use is greatest among white and publicly insured children. However, the American Academy of Otolaryngology-Head and Neck Surgery's 2011 clinical practice guideline statement was associated with an immediate decrease and change in use trends in these groups, narrowing differences in utilization by race/ethnicity and type of insurance.

The hidden cost of deep neck space infections

Introduction The incidence of deep neck space infection (DNSI) is rising and appears to be related to falling rates of tonsillectomy. The purpose of this study was to assess demographics of patients presenting with DNSI and the financial burden to the National Health Service (NHS). Methods Data were collected retrospectively on patients aged over 16 years admitted to NHS Greater Glasgow and Clyde with DNSI between 2012 and 2016. Demographics, aetiology and use of hospital resources were reviewed. The cost of hospital admissions was calculated using data from NHS Scotland's Information Services Division, the local diagnostics division and the British National Formulary. Results Seventy-four patients were admitted with DNSI during the study period. Forty (54%) were male. The mean age was 44.0 years (range: 16-86 years). The most frequent source of infection was the tonsil (n=30, 40.5%). The most common infective organism was Streptococcus constellatus (n=9, 12.2%). The mean length of stay was 11 days. Fifty-five patients (74.3%) required operative intervention. The mean cost of admission per patient was £5,700 (range: £332-£46,700). Conclusions This study highlights the high cost burden of DNSI to the NHS. The incidence of DNSI in Glasgow has risen over the study period; contributing factors may include the reduced tonsillectomy rate and a reduction in antibiotic prescribing. As the incidence of DNSI continues to rise, there will be an increase in cost to the NHS, which must be planned for.

Emergency department visits, hospitalizations, and readmissions of patients with a peritonsillar abscess.

To determine the demographics and treatment outcomes of patients with a peritonsillar abscess (PTA) and to provide guidance for treatment options. The 2012 National Emergency Department Database, 2012 National Inpatient Sample, and 2013 Nationwide Readmissions Database were used to analyze patients with a PTA who presented to emergency departments (ED), were admitted and/or readmitted to hospitals in the United States. Outcomes were used to create a decision model to compare initial medical versus surgical management. An estimated 62,787 ED visits; 15,095 inpatient admissions; and 267 readmissions for PTA were recorded. The average age was 29 years old, with a slight male predominance. The majority of patients seen in the ED (80%) were treated without surgery and discharged home. Patients admitted to hospital were more likely to undergo surgery (50%). The tonsillectomy rate was 11%. The complication and readmission rates were < 2%. Medical therapy was used more often than surgical therapy in many cases. Peritonsillar abscess mostly occurs in adults and leads to a significant number of ED visits, admissions, and readmissions. A trial of medical management of PTA appears to be safe, although surgery is highly successful and associated with low morbidity. 2c. Laryngoscope, 127:S1-S9, 2017.

Tonsillectomy rates in the South African private healthcare sector.

Adeno-/tonsillectomy is a commonly performed procedure with internationally standardised and recognised indications. Despite this, there exists considerable international (190 - 850/100 000 people ≤19 years of age) and regional variation in rates. This cannot be accounted for by differences in clinical need or regional morbidity. To describe the adeno-/tonsillectomy rate in the South African (SA) private healthcare sector and regional variations thereof. To compare local rates with international rates and assess trends in adeno-/tonsillectomy practice. Analysis of 2012 and 2013 adeno-/tonsillectomy data provided by the largest SA private healthcare funder, accounting for 30% of the medical scheme market. Rates are expressed per 100 000 people ≤19 years of age. The tonsillectomy rate in the SA private healthcare sector was 1 888/100 000 people ≤19 years of age in 2012. In 2013, the rate dropped significantly (p&lt;0.001) to 1 755/100 000. This is more than double the highest national tonsillectomy rate reported in the literature. There was also considerable regional variation in this rate within SA. The SA tonsillectomy rate is very high when compared with international trends and varies regionally within the country. The literature does not support an increased burden of disease as the reason behind this. Rather, it is differences in training and clinical practice of clinicians, as well as social and family factors, that have been implicated. The adeno-/tonsillectomy rate in the SA private healthcare sector is substantially higher than international norms. The reasons for this discrepancy require further consideration.

Hospital discharge survey on 4,199 peritonsillar abscesses in the Veneto region: what is the risk of recurrence and complications without tonsillectomy?

The Veneto region's database of hospital discharge records was queried for ICD-9 codes corresponding to: peritonsillar abscess (PTA), PTA incision and drainage, tonsillectomy, pharyngeal-retropharyngeal abscess, cervical phlegmon, cervical abscess, and mediastinitis recorded from 1997 to 2006. All these codes were considered to identify cases of PTA recurrence and severe infectious complications occurring in conservatively treated patients. Among 4,199 patients whose PTA was incised and drained on admission to hospital, 1,532 were treated with tonsillectomy, while 2,667 were treated conservatively (without tonsillectomy). Abscess tonsillectomy was carried out almost exclusively in children (0-14 years of age), and only in 40 young and adult patients (0.95 %). The relapse rate after a single episode of PTA was 11.7 %, while potentially fatal complication occurred in 0.41 % of cases. Incidence of PTA hospital admission has remained stable in the considered period despite a 45 % reduction in the tonsillectomy rate. In conclusion, our data seem to show that conservative treatment for PTA is not associated with a significant risk of recurrence (and becomes minimal after 6-12 months), provided that patients have not suffered from previous PTA episodes.

The impact of prior tonsillitis and treatment modality on the recurrence of peritonsillar abscess: a nationwide cohort study.

BackgroundStudies suggest an increased risk of peritonsillar abscess (PTA) recurrence in patients with prior tonsillitis. However, this association is inconsistent and could be confounded by different treatment modalities. This study aimed to assess the risk of recurrence among PTA patients with different degrees of prior tonsillitis and treatment modalities, and the role of tonsillectomy in current practice.MethodsAll in-patients with peritonsillar abscess between January 2001 and December 2009 were identified in a nationwide, retrospective cohort study. Recurrence was defined as the first occurrence of PTA ≧30 days from the initial PTA. Factors independently associated with recurrence were analyzed using Cox proportional hazard model after adjusting for demographic and clinical data.ResultsThere were 28,837 patients, with a 5.15% recurrence rate and 4.74 years of follow-up. The recurrence rates were significantly higher among subjects with more than five prior tonsillitis or 1–4 prior tonsillitis compared to those without prior tonsillitis (adjusted hazard ratio, 2.82 [95% confidence interval, 2.39–3.33] and 1.59 [95% CI: 1.38–1.82]). The adjusted HR in patients treated with needle aspiration was 1.08 compared to those treated with incision & drainage (95% CI: 0.85–1.38). After age stratification, the adjusted HRs of more than five prior tonsillitis increased to 2.92 and 3.50 in patients aged ≦18 and 19–29 years respectively. The adjusted HR ofneedle aspiration only increased in patients ≦18 years old (aHR: 1.98 [95% CI: 0.99–3.97]). The overall tonsillectomy rate was 1.48% during our study period.ConclusionsThe risk of PTA recurrence increases with higher degrees of prior tonsillitis in all age groups and management by needle aspiration only in the pediatric population. Patients younger than 30 years old with PTA and more than five prior tonsillitis have the greatest risk of recurrence.

The rising rate of admissions for tonsillitis and neck space abscesses in England, 1991-2011.

Sore throats and tonsillitis represent a considerable health burden as well as a significant source of expenditure for the National Health Service (NHS). As part of the recent NHS savings drive, the introduction of 'procedures of low clinical effectiveness' (PoLCE) lists has reinforced a large reduction in the number of tonsillectomies performed. We carried out a cross-sectional study of trends in emergency sore throat admissions in the context of the number of tonsillectomies performed. Hospital Episode Statistics (HES) data were extracted. Office for National Statistics data were also used. Between 1991 and 2011, the overall tonsillectomy rate fell by 44%. In the same time, the admission rate for tonsillitis rose by 310% (Pearson's r=-0.67, p=0.01). The peritonsillar abscess admission rate rose by 31% (r=-0.79, p<0.01). Between 1996 and 2011, the overall tonsillectomy rate fell by 41% and the retro and parapharyngeal abscess admission rate rose by 39% (r=-0.55, p=0.026). There was a 14% overall increase in tonsillectomy and sore throat associated bed days. This was despite the large fall in tonsillectomy numbers and the reduction in length of hospital stay. Efforts to reduce the tonsillectomy rate are correlated with a significant rise in emergency admissions. The rise in the retro and parapharyngeal abscess rate is perhaps most alarming given the very high mortality of these conditions. Bed day data suggest that no net saving has been made despite the new measures.

Laser tonsil cryptolysis: In-office 500 cases review

ObjectivesTonsilloliths, proven to be tonsillar biofilms cause symptoms of halitosis, foreign body sensation and recurrent sore throats. Laser Tonsil Cryptolysis (LTC) performed in the office may represent an alternative to tonsillectomy in selected cases of persistent tonsilloliths with cryptic infections. Study DesignA retrospective chart analysis using CPT codes. SettingOffice and hospital. MethodsA retrospective complications review consisting of bleeding, the need for an additional procedure, patient satisfaction and conversion rate to complete tonsillectomy was documented. ResultsFive hundred consecutive LTCs performed in the office under local anesthesia with a CO2 or diode laser were identified. Energy delivery was in continuous mode with power settings of 18W and 10W respectively. Bleeding occurred in 6 patients requiring unscheduled return office visit for evaluation. Eighty patients required a second procedure, comprising total of 1.16 procedures per patient. Eighteen (3.6%) patients underwent complete tonsillectomy. Patient satisfaction was high with an overall incidence of 0–2 days of work absence. Follow-up was 1–8 years. ConclusionsWith a small tonsil size, controllable gag reflex and cooperative adult patient LTC allows several advantages compared to conventional tonsillectomy. Benefits of LTC include avoidance of general anesthesia and limited ablation of cryptic pockets, resulting in reduced post-operative pain, bleeding, shorter recovery time and the convenience and cost advantage of an office procedure. With 1.16 sessions required per patient, low conversion rate to standard tonsillectomy and minimal complication rate LTC can be considered an alternative option to a patient suffering from recurrent cryptic tonsillitis with or without tonsilloliths.

Variation in practice: an analysis of Scottish Surgical Profiles ENT data

Variation in otolaryngology intervention rates is reported in the Scottish Surgical Profiles Project. Tonsillectomy is one of the selected key indicator procedures. The variation in practice was discussed nationally at the Scottish Otolaryngology Society summer meetings in 2009 and 2010. NHS Grampian had a significantly higher tonsillectomy rate compared with other Scottish NHS boards. To determine the accuracy of NHS Grampian data reported by the Information Service Division (ISD) and to record the appropriateness of listing of patients for tonsillectomy with reference to the Scottish Intercollegiate Guidelines Network (SIGN). Retrospective review of case notes and surgical records of patients who had undergone tonsillectomy between March 2007 and March 2008 in NHS Grampian. Between March 2007 and March 2008, 509 tonsillectomy cases were performed in NHS Grampian. This corresponded to the data received from ISD. 87% of tonsillectomies performed were compliant with SIGN guidelines. The Scottish otolaryngology clinicians have found the reporting of the intervention rates stimulating and challenging. Discussion of the surgical profile project regularly at national specialty meetings resulted in a preliminary detailed targeted audit of those who were persistent outliers for tonsillectomy. This refuted the presumed reasons for this variation, namely inaccurate figures from ISD and inappropriate listings by clinicians.

Exploring the variation in paediatric tonsillectomy rates between English regions: a 5‐year NHS and independent sector data analysis

Tonsillectomy rates vary between locations and socio-economic groups and have been noted to do so since the 1930s. We aimed, first, to illustrate the current extent of variation in rates in the NHS across England and, second, to explore whether any observed geographical variation could be accounted for by combining independent sector data with standard NHS datasets. Analysis of NHS Hospital Episode Statistics data for tonsillectomy in people aged under 15 years, from April 1 2000 to March 31 2005; addition of data from a major independent sector provider of surgery; ranking of local authority areas according to tonsillectomy rates including and excluding independent sector data. The national annual average tonsillectomy rate for the NHS was 304 per 100 000 population aged <15 (95% CI 320-324). The highest rate was 754 (95% CI 690-822) and the lowest 102 (95% CI 83-125). This seven-fold difference cannot simply be explained by a small number of high or low 'outliers'. When rates in the NHS were ranked and compared with the ranking of rates in the independent sector data, a Kendall tau rank correlation coefficient of 0.003 showed that there was no inverse correlation between the two rankings. That is, low NHS rates were not associated with high independent sector rates or vice versa. There is currently a seven-fold variation in tonsillectomy rates across local authority areas in England. The variation in tonsillectomy rates found in NHS data on surgical workload does not appear to be explained by the exclusion of data from the independent sector.

Day-case tonsillectomy in children: parental attitudes and consultation rates

Objective: In Europe, the day-case tonsillectomy rate in children is slowly increasing, but whether parents really want this rapid discharge of their child is questionable. The fear is that aftercare might fall solely on community care. The aim of this prospective study was to introduce pediatric day-case tonsillectomy to our hospital and to determine parents’ attitudes to this procedure. The other interest centered on consultation rates within the 2-week recovery period. Methods: One hundred children aged 3–16 years had day-case tonsillectomy (38) or adenotonsillectomy (62). Peri-operative and post-operative complications were recorded. Parents were phoned the next day and 1–4 months after the operation. Parents’ opinions of day-case surgery and consultations with healthcare professionals during the 2-week recovery period were recorded. Results: Ninety children went home the day of the operation. Vomiting was the most frequent complication. No primary hemorrhages occurred. Called the next day, 100% of parents felt that their children were better served spending their first night at home as compared with staying in hospital. Called 1–4 months later, 94.5% of parents still thought this way. In the 2-week recovery period following the tonsillectomy, 13% of patients visited a physician and 17% called for information. These numbers include patients with secondary hemorrhage. If these are excluded, 5% of patients visited a physician and 13% called for advice. Children were taken back to hospital only due to secondary hemorrhage. Conclusions: Most parents considered day-case tonsillectomy to be suitable for their family. Consultation rates were low. Careful patient selection and adequate pre-operative information are prerequisites for day-case tonsillectomy.

The management of severe infectious mononucleosis tonsillitis and upper airway obstruction.

Acute tonsillectomy has been advocated for severe infectious mononucleosis (IM) pharyngotonsillitis with upper airway obstruction (UAO) and not responding to corticosteroids. This paper reviews and rationalizes our management. A five-year chart review was carried out with a follow-up telephone survey. There were 36 admissions. Nine had UAO of whom 89 per cent (eight of nine patients) had a tonsillectomy. The diagnosis of UAO was not standardized. Twenty-seven patients did not have UAO and only a small proportion of these patients, 16 per cent (four of 25) went on to have an elective tonsillectomy. Corticosteroids made no significant difference to the tonsillectomy rate. UAO appears to identify patients with more severe disease who are likely to suffer recurrent tonsillitis. Acute tonsillectomy is an appropriate treatment option for this subgroup. The diagnosis of significant IM UAO is not defined and a schema is proposed. Recurrent tonsillitis is an uncommon sequela of severe IM pharyngotonsillitis without UAO.

Tonsillectomy and inflammatory bowel disease location.

Epidemiological studies have found an increased frequency of childhood infections and tonsillectomies in patients with inflammatory bowel disease (IBD). The aim of our study was to test whether particular clinical patterns of IBD could be associated with previous tonsillectomy. Two hundred and twenty consecutive IBD patients (100 with Crohn's disease (CD) and 120 with ulcerative colitis (UC)) were prospectively assessed and classified into groups according to disease location. In those with a positive history of tonsillectomy, an ear, nose and throat (ENT) examination was carried out to confirm the diagnosis. Eighty-two IBD patients (37%) underwent tonsillectomy in childhood: 47 out of 100 CD patients (47%), and 35 out of 120 UC patients (29%). Disease location in CD patients affected the ileum in 28, and 21 of them were tonsillectomized (tonsillectomy rate of 75%). Thirty-six patients had an ileocolonic location, and 14 of them had had a tonsillectomy (39%). With disease confined to the colon, only 9 out of 29 patients (31%) had undergone previous tonsillectomy. The higher prevalence of tonsillectomy in patients with CD ileitis was statistically significant (P = 0.0034). No significant differences between groups of UC patients according to the extent of the disease were found. The ileum is the most prevalent location of disease in CD patients with previous tonsillectomy.

Analysis of secular trends in surgery for glue ear in the North Western Region (1975-1984).

Hospital Activity Analysis (HAA) data relating to operations for glue ear and tonsillectomy performed on residents of the North Western Region, aged 0-9 years during the period 1975-1984 were studied. The rate of surgery for glue ear has risen from 45/10,000 to 107/10,000 representing an increase of 137 per cent. The tonsillectomy rate by comparison has undergone a more modest increase of 19 per cent. The study casts doubt on a 'vacuum effect', as has been argued by Black. Marked variation in rates of surgery between Districts in the Region were noted, and the possible explanations discussed. A review of the literature revealed considerable controversy concerning the management of glue ear and the need for a definitive trial is highlighted.

Geographical Variations in Use of Surgery for Glue Ear

Study of geographical variations in the rate of surgery for glue ear reveals striking differences both between English health regions and between health districts. There are two-fold differences between regions and up to seven-fold differences between districts. Analysis of these differences at different levels of population aggregation reveals that professional uncertainty about the indications and value of surgery for this condition is the major factor responsible. In addition, the availability of otolaryngologists influences the surgical rate at district though not at regional level. The method used in this study is unable to determine the part played by geographical variation in patient expectation and morbidity rate. In contrast to the lack of consensus among otolaryngologists regarding the use of surgery for glue ear, there appears to be a high degree of agreement as regards tonsillectomy. The latter has coincided with the end of the decline in the tonsillectomy rate in the late 1970s.

Tonsillectomy history in Hodgkin's disease.

Abstract Of 174 patients recently treated for Hodgkin's disease, a history of tonsillectomy was obtained in 52 per cent. Except for a low rate of tonsillectomy in younger patients, statistically significant variations from the overall rate were absent when tonsillectomy was correlated with other features such as sex, extent of disease, presence versus absence of constitutional symptoms, primary site of involvement and histologic subclassification. A comparison of patients and matched siblings did not reveal a difference in tonsillectomy rate although a number of patients with intact tonsils had healthy matched siblings with a history of tonsillectomy. The review failed to demonstrate a relation between tonsillectomy and the clinical features of Hodgkin's disease and did not support the hypothesis that prior tonsillectomy increases susceptibility to this disease by removal of a lymphoid barrier. Also of interest were two sibling pairs without tonsillectomy in whom Hodgkin's disease developed coincidental...

INCIDENCE OF CRANIAL NERVE PARALYSIS IN POLIOMYELITIS IN RELATION TO PRESENCE OR ABSENCE OF TONSILS

A study of patients from a largely rural area in Iowa corroborates the evidence of many previous studies that bulbar and bulbospinal types of poliomyelitis occur more commonly in persons whose tonsils have been removed irrespective of the time in life the operation was performed. The tonsillectomy rate for all cases was the lowest encountered thus far. Agreement is not on the same level as noted in the author's Detroit composite study previously reported, and is more marked for bulbospinal than for the bulbar type of poliomyelitis (some studies combine bulbar and bulbospinal cases). Adjustment for age reduces differences between cases with tonsils absent and tonsils present by clinical type and further strengthens the finding of Paffenbarger and of the author's Detroit composite study that age must be reckoned with in an assessment of the problem. Incidence of paralyses of cranial nerves by tonsillectomy status is similar to the Detroit composite study in the case of palatal and pharyngeal paralyses but less strongly; incidence of paralysis of the facial nerve is dissimilar.

A study of rheumatic fever and streptococcal infection in different social groups in Melbourne.

1. Hospital records of 1469 cases of rheumatic fever (including chorea) up to the age of 14 years were used to study the social distribution of the disease in Melbourne. On this evidence the incidence of rheumatic fever was three times greater in low than in high rental districts.2. The results of throat-swab surveys and Dick tests on schoolchildren living in different districts showed that children in the poor districts contracted streptococcal infections more frequently than those in well-to-do districts.3. One factor associated with the carrier rate ofStreptococcus pyogenesin the various social groups was the incidence of tonsillectomy. In the better class districts where the tonsillectomy rate was high (60%) theStr. pyogenescarrier rate was 12.5%, whereas the rates in poor areas were 35 and 25.3% respectively.4. Differences in social incidence of rheumatic fever might therefore be explained by differences in social incidence of streptococcal infection, which, in turn, might be influenced by the incidence of tonsillectomy.We wish to thank the many persons who so willingly assisted in the collection of data for this study, particularly Dr H. L. Stokes and Dr M. M. Wilson, and Dr R. E. O. Williams for advice on presentation. This work was supported by a grant to one of us (Dr M. C. Holmes) from the Commonwealth Research Fund from the University of Melbourne.