TECHNOLOGY OVERVIEW. Issue 19 March 2006

Transcription

1 TECHNOLOGY OVERVIEW Issue 19 March 2006 Clinical and Cost-effectiveness of Screening Newborns for Medium Chain Acyl~CoA Dehydrogenase Deficiency Using Tandem Mass Spectrometry

2 Publications can be requested from: CCOHTA Carling Avenue Ottawa ON Canada K1S 5S8 Tel. (613) Fax. (613) or download from CCOHTA s web site: Cite as: Tran K, Banerjee S, Li H, Noorani H, Mensinkai S, Dooley K. Newborn screening for medium chain acyl~coa dehydrogenase deficiency using tandem mass spectrometry [Technology overview no 19]. Ottawa: Canadian Coordinating Office for Health Technology Assessment; This report and the French version entitled Étude de l'efficacité clinique et du rapport coût-efficacité du dépistage néonatal du déficit en acyl-coenzyme A déshydrogénase des acides gras à chaîne moyenne par le biais de la spectrométrie de masse en tandem are available on CCOHTA s web site. Production of this report is made possible by financial contributions from Health Canada and the governments of Alberta, British Columbia, Manitoba, New Brunswick, Newfoundland and Labrador, Northwest Territories, Nova Scotia, Nunavut, Ontario, Prince Edward Island, Québec, Saskatchewan, and Yukon. The Canadian Coordinating Office for Health Technology Assessment takes sole responsibility for the final form and content of this report. The views expressed herein do not necessarily represent the views of Health Canada or any provincial or territorial government. Reproduction of this document for non-commercial purposes is permitted provided appropriate credit is given to CCOHTA. CCOHTA is funded by Canadian federal, provincial and territorial governments. Legal Deposit 2006 National Library of Canada ISSN: (print) ISSN: (online) PUBLICATIONS MAIL AGREEMENT NO RETURN UNDELIVERABLE CANADIAN ADDRESSES TO CANADIAN COORDINATING OFFICE FOR HEALTH TECHNOLOGY ASSESSMENT CARLING AVENUE OTTAWA ON K1S 5S8

3 The Canadian Coordinating Office for Health Technology Assessment Clinical and Cost-effectiveness of Screening Newborns for Medium Chain Acyl~CoA Dehydrogenase Deficiency Using Tandem Mass Spectrometry March 2006 We thank Suzanne Morphet for her assistance in creating this overview from a longer report authored by Khai Tran et al. This Overview is based on the Technology Report commissioned by CCOHTA: Tran K, Banerjee S, Li H, Noorani H, Mensinkai S, Dooley K. Newborn screening for medium chain acyl~coa dehydrogenase deficiency using tandem mass spectrometry: clinical and cost-effectiveness [Technology report no 62]. Ottawa: Canadian Coordinating Office for Health Technology Assessment; CCOHTA takes sole responsibility for the final form and content.

4 REPORT IN BRIEF March 2006 Newborn Screening for Medium Chain Acyl~CoA Dehydrogenase Deficiency Using Tandem Mass Spectrometry: Clinical and Cost-effectiveness Technology Tandem mass spectrometry (MS/MS)-based newborn screening for medium chain acyl~coa dehydrogenase deficiency (MCADD). Disease MCADD is an inherited metabolic disorder. In Canada, there are about 20 incident cases/year. Most MCADD patients are asymptomatic before an acute episode, which manifests with signs of encephalopathy, hypoglycemia, or sudden death. Of those who have an acute episode, almost onequarter will die, and one-third of survivors will have irreversible neurological damage. Issue Newborn screening for certain inborn errors of metabolism is standard across Canada, but screening for MCADD is not consistently practised. Thus, many cases of MCADD are diagnosed clinically, often during an acute episode. Because it is treatable if diagnosed early, there is a need to determine whether jurisdictional newborn screening programs should be expanded to include MCADD. Methods and Results A systematic review of the clinical and economic literature was performed. The clinical sensitivity and specificity of MS/MS-based screening for MCADD were very high. The economic review and analysis showed that screening results in more quality-adjusted life-years (QALY), and lower morbidity and mortality compared with no screening. The economic review showed that screening is cost-effective compared to no screening if willingess to pay is $50,000 per QALY. The primary economic analysis using Canadian data showed that screening is cost-effective if willingness to pay is $20,000 per QALY. Implications for Decision Making Newborn screening of MCADD by MS/MS identifies patients while they are asymptomatic. Without screening, diagnoses of MCADD are typically made at clinical presentation, often during an acute episode. MS/MS screening for MCADD has high clinical validity. The addition of MCADD to Canadian newborn screening programs would identify about 20 cases/year. MS/MS-based screening for MCADD consumes more resources but attains better health outcomes than no screening. Based on a primary economic analysis, screening is cost-effective if willingness to pay is $20,000 per QALY. Ethical and psychosocial issues, such as informed consent, false diagnosis, parental anxiety, privacy, and confidentiality, warrant consideration. This summary is based on a comprehensive health technology assessment available from CCOHTA s web site ( Tran K, Banerjee S, Li H, Noorani HZ, Mensinkai S, Dooley K. Newborn screening for medium chain acyl~coa dehydrogenase deficiency using tandem mass spectrometry: clinical and cost-effectiveness. Canadian Coordinating Office for Health Technology Assessment (CCOHTA) Carling Avenue, Ottawa ON Canada K1S 5S8 Tel: Fax: CCOHTA is an independent, not-for-profit organization that supports informed health care decision making by providing unbiased, reliable information about health technologies..

5 1 Introduction Medium chain acyl~coa dehydrogenase deficiency (MCADD) is a rare but potentially fatal genetic disorder that can be prevented by early diagnosis and treatment. It is an autosomal recessive inherited abnormality, and one of the most common metabolic defects of fatty acid ß- oxidation. When energy is needed during fasting or exercise, the body releases fatty acids from adipose tissue. The fatty acids are taken up by cells in the heart, muscle, and liver, and converted to acyl~coa esters in the cytoplasm. 1 The acyl~coa esters cross the mitochondria membranes as acylcarnitines, then return to their original form, and undergo oxidation, which is mediated by long, medium, and short chain acyl~coa dehydrogenases. If there is a defect in one of the enzymes and the acyl chain is incompletely oxidized, it accumulates in the plasma as acylcarnitines. As a result, MCADD patients have higher than normal levels of octanoylcarnitine (AC8) in their plasma, or an increased ratio of AC8 to decanoylcarnitine (AC10). Approximately 30 mutations have been found in the MCADD gene. 2 More than 80% of clinically diagnosed patients are homozygous for the A985G mutation, and 18% are compound heterozygous. 3 About half of the patients who are detected by newborn screening are homozygous for the A985G mutation. Another mutation, T199C, also causes MCADD. While it has not been found in clinically diagnosed patients, it is present in a large proportion of samples from newborn screening. 4 The incidence of MCADD is higher in Germany, the UK, and English-speaking countries, because the common A985G mutation originated in people of northern European descent. 5,6 Each year in the UK, between 35 and 70 newborns have MCADD, 7 an incidence of 1:20,000 to 1:9,091; in Germany, the incidence is 1:9,773. In Canada, given our multicultural population and current birth rate, the incidence is estimated to be 1:16,000, or about 20 cases annually. 8 Most often, the symptoms of MCADD occur within the first two months of life, with signs of acute encephalopathy and hypoglycemia. Up to a quarter of patients die, and one-third of the survivors have irreversible brain damage and may require lifetime hospital care. 3,9,10 Timely intervention, monitoring by a metabolic disease clinic, and a high carbohydrate diet during fasting or gastrointestinal infection seem to be effective in preventing recurring episodes in affected individuals. 9,11 Because newborn screening is new, it is unknown how the patients who are identified by screening fare in the long term. The screening method of choice for MCADD involves two mass spectrometers in tandem. The first separates compounds from a blood sample by their molecular mass. The resulting ions are then put through a collision cell, where they are bombarded by gas and turned into product ions. These ions are passed to a second spectrometer for an analysis of their masses, so that the quantities of the molecule can be determined. Abnormal levels of the analytes of interest can be detected based on a predetermined cut-off concentration. Tandem mass spectrometry (MS/MS) has high specificity and sensitivity; appropriate cut-off concentrations of AC8 must be determined to avoid excessive false-positive results. Most 1

6 screening programs use a cut off of 0.5 µm to 1.0 µm. 12,13 Because serum concentrations of AC8 decrease as time progresses after birth, blood samples should be collected before 72 hours of age to help avoid false-negative results. A molar ratio of AC8 to AC10 is used as an additional identifier. 12 Positive results should be followed up with molecular testing, determination of fibroblast enzyme activity, and a clinical evaluation to confirm or rule out the diagnosis. Screening facilities for MCADD can be established in hospitals or public health laboratories that offer access to clinical geneticists, genetic counsellors, dietitians, and a specialized metabolic clinic. In Canada, these facilities are available in most provinces, but they may be inadequate for MCADD testing. 2 Objective The objective is to review the potential application of MS/MS to detect MCADD in newborns, in the context of Canadian screening programs, taking into account clinical, financial, ethical, and psychosocial issues. This report will help policy makers and program managers decide whether to include MCADD in existing screening programs. The questions that will be addressed by the report include: What is the evidence of benefit, harm, and validity (sensitivity, specificity, and predictive values) in screening newborns for MCADD using MS/MS? What are the clinical outcomes of patients with MCADD whose conditions were detected at birth by screening compared with those who were diagnosed later in childhood? What is the cost-effectiveness of screening for MCADD using MS/MS? What is the budget impact of screening for MCADD using MS/MS? What are the ethical and psychosocial issues associated with newborn screening for MCADD using MS/MS? 3 Clinical Review Methods A systematic review with a protocol written a priori was followed. For the clinical review, we found published literature by cross-searching MEDLINE, BIOSIS Previews, PASCAL, Social SciSearch, PSYCInfo, ERIC, and EMBASE databases from 1995 onwards. We also ran parallel searches on PubMed, CINAHL, and Cochrane databases. We obtained grey literature by searching specialized databases, and web sites of regulatory agencies and health technology related agencies. Selection Criteria To be included in the report, studies had to specify newborn screening for MCADD using MS/MS; compare outcomes of MS/MS-based screening and clinical diagnosis; or report outcomes of MCADD patients who were diagnosed clinically. 2

7 Two reviewers independently screened all titles and abstracts, after which they reviewed the full text articles, applying the selection criteria. Differences about which studies to include were resolved by consensus, and the content expert author verified all decisions. Quality Assessment Two reviewers used QUADAS, 14 a quality assessment tool for studies of diagnostic accuracy included in systematic reviews. QUADAS consists of 14 questions that cover bias, variability, and reporting. Data Analysis Data for MS/MS-based screening included incidence, rate of detection, sensitivity, specificity, and predictive values. Markers for MCADD were AC8 levels, and the ratio of AC8 to AC10. The DNA analysis identified homozygosity or compound heterozygosity for the common A985G mutation, for which the percentage could be calculated. For clinically diagnosed patients, data included rate of detection; numbers of symptomatic patients, asymptomatic patients, heterozygotes, and homozygotes; and outcomes such as hypoglycemia and lethargy, coma and encephalopathy, neurological impairment and developmental delay, full recovery, and death. We calculated the 95% confidence interval (CI) and weighted means for these data (Appendix 6 of the Technology Report). 15 Results We found 957 citations in our literature search, of which we identified 48 reports as potentially relevant. After retrieving the full text of these reports, 21 were selected for data extraction. 3,4,12,16-33 Of these, 15 were reported as full-length articles, 3,4,12,16-21,31-33 and six were reported as abstracts. 22,26-33 The 27 excluded articles were eliminated, because they lacked specific data about MCADD; duplicated some included studies; focused on establishing test criteria; or had an inappropriate design. We assessed the quality of 14 of the 21 included studies using the QUADAS scale. Most were of limited quality. For instance, in 12 of the 14 studies, the populations were specific ethnic groups. This suggests that the results may have limited external validity. There was often no indication of the length of time between positive screening results and a confirmation of those results. Of the seven studies that were not assessed for quality, six were reported in abstract form, and one was a retrospective evaluation of the outcomes of clinically diagnosed MCADD patients. Trial Characteristics Of the 21 studies, four compared outcomes of patients identified by MS/MS versus clinical diagnosis, 17,19,21,23 one studied patient outcomes after diagnosis, 3 and 16 presented the results of patients identified only by MS/MS screening. 4,12,16,18,20,22,24-33 All but one were prospective studies; the exception 24 used MS/MS to screen newborn blood samples that had been collected in the two and a half years before the study, and compared the results with those obtained by clinical diagnosis. The study period varied from one to eight years for MS/MS-based screening, and one to 24 years for clinical diagnosis. The number of newborns screened ranged from 9,320 to 930,078. The sampling period varied, but usually occurred after the second or third day of life. 3

8 The age of patients who were clinically diagnosed ranged from two days old to 10.5 years old. The AC8 cut-off levels shown in 10 studies 4,12,16,20,21,23-25,28,32 varied from 0.25 µm to 1 µm. Three studies 12,18,28 appear to have been funded by Neo Gen Screening Inc. Data Analysis and Synthesis The calculated incidence of MCADD, based on the detected cases and study population, varied from 1:68,560 to 1:8,930, The mean was 1:16,667 (95% CI 1:20,000; 1:14,286). The rate of detection varied from 0.25 to 17 cases annually, with a mean of 5.4 (95% CI: 4.3 to 6.4). The highest incidence and rate of detection were reported in two studies in Germany, 16,17 but a high incidence was also found in northern UK (1:12,600). 24 Information about plasma AC8 levels, ratios of AC8 and AC10, and the percentage of A985G mutation in individuals identified by MS/MS screening is presented in Table 1. Few of the studies that looked at MS/MS-based screening followed up or reported on patient outcomes, compared with studies of those who were clinically diagnosed in childhood. Almost all MCADD patients identified by MS/MS were asymptomatic whereas their clinical counterparts usually had symptoms. Treatment was generally avoidance of fasting, diets low in fat and high in carbohydrate, and supplementation with carnitine. Those MS/MS-screened patients who died usually experienced complications at the start of screening; a mean 2% of screened patients died (95% CI: 1 to 6), compared with 16% of patients identified by clinical diagnosis (95% CI: 8 to 26). There were also differences in the percentage of A985G mutation homozygotes and heterozygotes who were detected by screening versus clinical diagnosis, with screening finding more compound heterozygotes, and clinical diagnosis finding mostly homozygotes. 4 Economic Analysis Methods As with the clinical review, we cross-searched MEDLINE, BIOSIS Previews, PASCAL, and EMBASE databases from 1995 onward, but this analysis was broadened to include descriptors for Inborn Errors of Metabolism. We used an economic filter to restrict findings to relevant economic studies. We also ran a parallel search on PubMed, CINAHL, and Cochrane databases; searched HEED: Health Economic Evaluations Database; and searched the web sites of the Canadian Institute for Health Information and the Ontario Case Costing Initiative. For grey literature, we searched the web sites of regulatory agencies, health technology assessment and related agencies and specialized databases, including the University of York NHS Centre for Reviews and Dissemination, and the Latin American and Caribbean Center on Health Sciences Information. 4

9 Table 1: Results of MS/MS analysis of acylcarnitine profile and DNA analysis for common A985G mutation Author AC8 (µm) AC8/AC10 Ratio Per Cent of Cases (country) Homozygous Compound Heterozygous Homozygous Compound Heterozygous Homozygous Compound Heterozygous Marsden (US) (3/7) 57 (4/7) *McCandless (US) (95% CI: 3.24 to 4.59) 1.86 (95% CI: 0.83 to 2.89) 10.2 (95% CI: 9.2 to 11.2) 6.0 (95% CI: 3.3 to 7.9) 71 (30/41) 25 (10/41) Carpenter (Aurtralia) 23 Screening: 5.6 (range 2.6 to 8.4) Follow-up: 9.6 (range 8.0 to 24.6) Screening: 7.5 (range 1.1 to 24.6) Follow-up: 2.1 (range 0.7 to 3.8) (4/12) 67 (8/12) Andresen (US) 4 >2 0.5 to 2.0 >4 2 to 4 63 (39/62) 34 (21/62) Zytkovicz 13.8 (range (range (4/10) 60 (6/10) (US) 25 to 22) to 3.2) *Marsden 25 (1/4) 75 (3/4) (US) 27 *McCandless (14/20) 30 (6/20) (US) 28 Sander 4.15 (mean of 5.86 (of (15/29) 3.4 (1/29) (Germany) patients) patient) Chace (US) (9/16) 44 (7/16) Ziadeh (US) (4/9) 56 (5/9) Mean (CI, 95%) 6 (0.9, 11.6) 4 (0.4, 7.2) 11 (7.8, 14.4) 5 (3.0, 6.9) 54 (46, 61) 34 (28, 41) *indicates abstract; AC8=octanoylcarnitine; AC10=decanoylcarnotine; compound heterozygous=heterozygous for two MCADD mutations. Selection Criteria Studies had to be full economic evaluations (such as cost minimization analysis, or costeffectiveness analysis) of infants or children with MCADD who were diagnosed by MS/MSbased screening or not screened, with the primary outcome being reported as an aggregated or disaggregated estimate of cost-effectiveness. Two reviewers independently screened all abstracts, then agreed on which full texts should be ordered. A study was included only if both reviewers answered yes to all criteria. Next, the reviewers extracted data using a structured form, with any disagreement resolved by consensus. Quality Assessment The British Medical Journal s 35-item checklist was used to assess the quality of included studies. Originally designed as a guideline for authors, it is a commonly used tool for assessing the quality of economic evaluations. 34 The checklist has three parts: study design, data collection analysis, and interpretation of results. Details of the checklist and our results can be found in Appendix 11 of the Technology Report. 15 5

10 Results Out of 289 abstracts, we ordered 29 full-length articles that met the inclusion criteria and retrieved 27. Of those, two were acceptable. 35,36 The remainder were excluded because the study type or study population were irrelevant to our analysis. The two accepted studies were American, and used primary and secondary data to present results from a societal perspective. While their model structures differed, both studies arrived at the same overall conclusion: MS/MS-based screening for MCADD is cost-effective compared with no screening, if willingness to pay is US$50,000 per quality-adjusted life-year (QALY). 5 Primary Economic Analysis Method We built a decision-tree model to analyze the cost-effectiveness of screening compared with no screening, based on what is known about how MCADD progresses, the impact of newborn screening on the health care system, and the availability of relevant Canadian data. There are 16 pathways for screening, depending on test results (positive or negative), MCADD status (true- or false-positive), time of acute episode (early or late), and long-term health consequences (asymptomatic, acute complications only, mild or severe neurological impairment, death); and there are six pathways for no screening, depending on MCADD status (with or without) and long-term health consequences. Results are expressed in terms of incremental cost, incremental effectiveness, and ICER, where applicable. Parameter Values and Assumptions We performed a sensitivity analysis to explore the effect of key parameter values on the economics of screening. Clinical data came from this report s clinical review, and cost-related data from the Nova Scotia screening program were obtained. The follow-up protocol for monitoring the health of MCADD patients also came from that program (Dr. Kent Dooley, Dalhousie University: unpublished observations, 2005). Details about the probabilities of adverse health outcomes, and calculations of the values of some cost-related parameters and assumptions for the base case scenario are presented in Appendices 12, 13, and 14 of the Technology Report. 15 Sensitivity Analysis We determined best and worst scenarios, and performed one- and two-way sensitivity analyses to dispel uncertainty in the base-case findings. In the best scenario, the incidence of MCADD is as high as 1:14,286, with the positive predictive value rising to 41%; the probability of being asymptomatic is 98% for screening, and 14% for no screening; and the probability of death linked to MCADD is 1% for screening, and 26% for no screening. 6

11 In the worst scenario, using the same data sources, the incidence of MCADD is as low as 1:20,000, and the positive predictive value is 33%; the probability of being asymptomatic is 92% for screening, and 35% for no screening; and the probability of death linked with MCADD is 6% for screening, and 8% for no screening. For the one-way sensitivity analysis, we changed the value of six parameters: specificity of screening from 99.95% to %; incidence from 1:20,000 to 1:14,286; screening cost from C$0.50 to C$5.60 per case; routine health monitoring cost from $1,500 to $4,000 per case; cost associated with an acute episode from $10,000 to $20,000 per episode; and direct medical cost due to severe neurological impairment from $100,000 to $250,000 per case during a lifetime. For a two-way sensitivity analysis, we varied four parameters, two at a time, simultaneously. These were based on the findings from the one-way sensitivity analysis. They were screening cost and specificity; and screening cost and cost associated with an acute episode. Results a) Base Case Screening consumed more resources but resulted in lower morbidity and lower mortality than no screening. The total cost for screening one cohort of 330,803 newborns during a 77-year horizon was $934, compared with $450, for no screening. The ICER was $2,514.02, which is below the threshold of C$20,000 suggested by Laupacis et al. 37 As a result, the base case results show that screening is cost-effective. b) Best and Worst Scenarios In best and worst scenarios, screening used more resources than no screening, but produced better health outcomes. The incremental cost was $389, in the best scenario, and $596, in the worst scenario, for screening versus no screening. The ICERs, expressed as incremental cost per QALY, were $ and $11, for best and worst scenarios respectively. c) One-way Sensitivity Analysis Varying the parameters produced different results. Increases in specificity, incidence rate, cost associated with acute episodes, and cost associated with severe neurological impairment pushed the ICER down. Conversely, increases in screening cost and management cost pushed the ICER up. Nonetheless, all ICER values were below the commonly used C$20,000 threshold for willingness to pay. d) Two-way Sensitivity Analysis When the ICER was calculated as a function of cost associated with specificity and screening, it ranged from -$756 to $7,875 per QALY. When it was calculated as a function of cost associated with an acute episode and screening, it ranged from -$860 to $7,363 per QALY. As a result, this analysis found screening to be cost-effective when willingness to pay is C$20,000 per QALY. 7

12 Limitations: In addition to the limited quality of the clinical studies discussed, we encountered limitations in our economic analyses. Neither of the economic evaluations that we considered included the QALY of family members, especially parents. The QALY was calculated using secondary data, rather than primary. Neither study reported on resources used or unit costs. Several potential limitations exist in our primary economic analysis. Uncertainty remains about the tested parameters because we obtained data from a provincial laboratory in Nova Scotia, and these data may differ from those of other provinces. Direct evidence about the long-term health outcomes of MCADD patients is unavailable. We based the value of the QALY on the values in three US studies, none of which had MCADD patients. The economic model is also potentially limiting, because the repetition of acute episodes, the cost difference of acute episodes occurring at different times, and the changing health outcomes were not considered. Also, a screening system that is devised to detect MCADD may not accommodate testing for other inborn errors of metabolism without additional capital costs, resource use, and health consequences. Health services impact: If all provinces and territories screen for MCADD, approximately 330,000 babies in Canada would be screened annually, and 20 would be found to have the disorder; another 33 would have false-positive results. A simple budgetary analysis based on a cohort of 8,533 newborns from Nova Scotia shows that in its first five years, a screening program would cost $317,921 ($220,419 in the first year alone), whereas no screening costs $121,841 for the same length of time. For screening, almost two-thirds of the budget would be spent in the first year, and most of that would go towards purchasing equipment. If no screening is done, the budget is spent on caring for patients with acute episodes and severe neurological impairment. This analysis is limited by the fact that key parameters, such as the cost of a more sophisticated MS/MS machine than the one budgeted for, are subject to change. Operational costs, including labour, may also differ by location, compared with those in the Nova Scotia screening program. Ethical and psychosocial issues: We systematically reviewed the literature for studies that addressed the ethical and psychosocial issues around screening newborns using MS/MS. We followed a similar search strategy as we did for the clinical and economic reviews, but did not restrict our search to MCADD alone; we looked at all relevant studies on inborn errors of metabolism. Out of 504 citations, we chose six articles reporting on three primary studies. 11,18,41-44 As with other genetic testing, screening newborns for disorders raises concerns about informed consent, false diagnosis, parental anxiety, privacy, and confidentiality of test results. With MCADD, even if the diagnosis is a true positive, 15% to 35% of patients will never experience ill effects, yet it is unknown which of these children will be asymptomatic. In general, testing for MCADD does not significantly influence the larger scale discussion of these concerns. As with any type of screening, parents should be adequately informed about the tests and the associated risks; they should also be given the option to explicitly refuse tests. 8

13 6 Conclusion Although the quality of the included studies was limited, the benefits of screening for MCADD using MS/MS-based technology outweigh the risks of no screening; it significantly reduced morbidity and mortality compared with clinical diagnosis. Screening using MS/MS-based technology costs more, but results in better health outcomes compared with no screening. Our economic review showed that MS/MS-based screening for MCADD is cost-effective, if willingness to pay is US$50,000 per QALY, while our primary economic analysis, which was based on Canadian data, showed the same but at a threshold of C$20,000 per QALY. 9

14 7 References 1. Roe C, et al. In: Scriver C, et al, editors. The metabolic basis of inherited disease. New York: McGraw Hill; p Andresen BS, et al. Hum Mol Genet 1997;6(5): Pollitt RJ, et al. Arch Dis Child 1998;79(2): Andresen BS, et al. Am J Hum Genet 2001;68(6): Wang SS, et al. Genet Med 1999;1(7): Liebl B, et al. Eur J Pediatr 2003;162 Suppl 1:S57-S Seddon HR, et al. Clin Chem 1997;43(3): Births and birth rate, by provinces and territories. Ottawa: Statistics Canada; Available: 9. Iafolla AK, et al. J Pediatr 1994;124(3): Wilcken B, et al. Arch Dis Child 1994;70(5): Wilson CJ, et al. Arch Dis Child 1999;80(5): Chace DH, et al. Clin Chem 1997;43(11): Van Hove JL, et al. Am J Hum Genet 1993;52(5): Whiting P, et al. BMC Med Res Methodol 2003;3(1): Tran K, et al. Newborn screening for medium chain acyl~coa dehydrogenase deficiency using tandem mass spectometry: clinical and cost-effectiveness. [Technology report no 62]. Ottawa: Canadian Coordinating Office for Health Technology Assessment; Sander S, et al. Eur J Pediatr 2001;160(5): Hoffmann GF, et al. Eur J Pediatr 2004;163(2): Marsden D. Southeast Asian J Trop Med Public Health 2003;34 Suppl 3: Wilcken B, et al. N Engl J Med 2003;348(23): Schulze A, et al. Pediatrics 2003;111(6 Pt 1): Shigematsu Y, et al. J Chromatogr B Analyt Technol Biomed Life Sci 2002;776(1): McCandless SE, et al. Am J Hum Genet 2002;71(4 Suppl): Carpenter K, et al. Arch Dis Child Fetal Neonatal Ed 2001;85(2):F105-F Pourfarzam M, et al. Lancet 2001;358(9287): Zytkovicz TH, et al. Clin Chem 2001;47(11): Wilcken B, et al. J Inherited Metab Dis 2000;23 Suppl 1: Marsden D, et al. J Inherited Metab Dis 2000;23 Suppl 1: McCandless SE, et al. Am J Hum Genet 2000;67(4 Suppl 2): Roscher A, et al. J Inherited Metab Dis 2000;23 Suppl 1: Abdenur JE, et al. J Inherited Metab Dis 2000;23 Suppl 1: Naylor EW, et al. J Child Neurol 1999;14 Suppl 1:S4-S Wiley V, et al. Acta Paediatr Suppl 1999;88(432): Ziadeh R, et al. Pediatr Res 1995;37(5): Drummond MF, et al. BMJ 1996;313(7052): Insinga RP, et al. J Pediatr 2002;141(4): Venditti LN, et al. Pediatrics 2003;112(5): Laupacis A, et al. CMAJ 1992;146(4): Neumann PJ, et al. Med Decis Making 2000;20(4): Erickson P, et al. Statistical Notes 1995;(7): Saigal S, et al. JAMA 1996;276(6): Waisbren SE, et al. JAMA 2003;290(19): Albers S, et al. J Inherited Metab Dis 2001;24(2): Liebl B, et al. Prev Med 2002;34(2): Liebl B, et al. Prev Med 2002;34(2):