Draft Genetic Test Review

---

Cystic Fibrosis
Clinical Utility

(347KB)

---

CLINICAL UTILITY

Question 26: What is the natural history of the disorder?
Question 27: What is the impact of a positive (or negative) test on patient care?
Question 28: If applicable, are diagnostic tests available?
Question 29: Is there an effective remedy or acceptable action, or other measurable benefit?
Question 30: Is there general access to that remedy or action?
Question 31: Is the test being offered to a socially vulnerable population?
Question 32: What quality assurance measures are in place?
Question 33: What are the results of pilot trials?
Question 34: What health risks can be identified for follow-up testing and/or intervention?
Question 35: What are the financial costs associated with testing?
Question 36: What are the economic benefits associated with actions resulting from testing?
Question 37: What facilities/personnel are available or easily put in place?
Question 38: What educational materials have been developed and validated, and which of these are available?
Question 39: Are there informed consent requirements?
Question 40: What methods exist for long term monitoring?
Question 41: What guidelines have been developed for evaluating program performance?

---

CLINICAL UTILITY

Question 35: What are the financial costs associated with testing?

page 1 | page 2 | page 3

---

Summary From the literature, eight cost components have been identified and assigned a reasonable unit cost estimate Providing education/information to the entire population                $1 to $3 Obtaining informed consent                                                        $5 to $10 Collecting and transporting the sample                                        $10 (blood) $4 (buccal) Performing the DNA test                                                             $80 to $100 Reporting negative results                                                          $2 (by mail/fax/electronic) Reporting positive results                                                            $20 (individual) $50 (couple) Performing diagnostic testing                                                       $400 to $600 (w/o karyotype) Accounting for procedure-related fetal losses                                $400   The costs per case of cystic fibrosis prenatally diagnosed are then computed using the above cost estimates, uptake rates, prevalences and proportions of mutations identified for each of five racial/ethnic groups. The costs per case detected using either the two-step (sequential) or the one-step (couple) model are $     500,000 for non-Hispanic Caucasians (the largest target population) $     400,000 for Ashkenazi Jewish $ 4,000,000 for Hispanic Caucasians> $ 7,000,000 for African Americans $19,000,000 for Asian Americans Using the expanded one-step (concurrent) model increases cost by about two-thirds. The costs per couple tested are essentially the same for various racial/ethnic groups, but are not the same for the various screening models  $120 to $140 using the two-step (sequential) or one-step (couple) model $220 to $240 using the expanded one-step (concurrent) model

Introduction
Economic costs and benefits can be categorized in multiple ways, varying with the purpose and perspective of the analysis. The simplest approach, used in most published evaluations of prenatal cystic fibrosis screening, is to analyze financial costs accruing to the health care system. The present analysis follows the health care system perspective. Only a few of the studies include more general societal costs (e.g., individual time and travel costs). A presupposition of the societal perspective is that the outcome of the intervention is socially valued as a good (e.g., prevention of early death). Since prenatal cystic fibrosis screening may lead to pregnancy termination, for which a social consensus is lacking, the societal perspective is not an appropriate choice. Analytic perspective also determines whether resource costs or charges should be assessed. The health care system perspective estimates the expenditure of real resources, adjusting for cost-shifting, economic rents, and transfer payments. Methods for estimating resource costs include use of internal accounting data, cost-to-charge ratios, or payments made by payers with strong market power. Each of these approaches may lead to an understatement of costs, either as a result of exclusion or undervaluation of costs or as a result of cost-shifting to other services. The alternative is to use stated charges or fees, which is appropriate if the analysis is being done from the perspective of a third-party payer with little market power. This review is restricted to direct financial costs and benefits to the medical care system and is divided into three parts.

1. Financial costs to the health care system include eight components:

2. Providing education/information to the entire population prior to offering the test

3. Obtaining informed consent from those choosing to be tested

4. Collecting and transporting the specimen (i.e., blood or buccal)

5. Performing the DNA test for CFTR mutations

6. Reporting negative results to the provider

7. Reporting positive results to the provider and patient with appropriate counseling

8. Performing diagnostic testing of the fetus (including counseling)

9. Accounting for fetal losses attributable to diagnostic testing

This section first presents published estimates of unit costs (or charges) for each of these eight components, followed by a brief discussion of each estimate and a summary conclusion. Once reasonable costs (and ranges) for each of seven components have been determined, a summary of the published assumptions relating to behavioral and epidemiologic parameters will be compared to results from actual pilot trials of prenatal cystic fibrosis screening. A final summary of the total cost of the screening process will then be provided. No two economic evaluations of prenatal cystic fibrosis screening evaluate the same protocol or protocols. Many compare one protocol with no screening for the purpose of evaluating whether screening makes economic sense. A few compare multiple protocols. Because even broadly similar protocols can be implemented in varying ways, it is difficult to develop generalizable cost estimates. All of the economic studies agree that the largest component of total cost is performing the DNA test (Component 4). The rapid evolution of molecular testing has continually altered both costs and the number of mutations that can be tested.

The published studies provide cost estimates in various currencies for currency years from 1993 to 1997. All estimates from other currencies are converted to US dollars at prevailing exchange rates. To standardize the estimates, costs are adjusted to 1996 dollars, using a relevant inflation index. For labor costs, this is the employment cost index and for other costs it is the GDP deflator for medical care. The laboratory cost of DNA testing is not adjusted for currency year, because laboratory costs likely reflect the rate of technological progress rather than general inflation. To adjust for inflation since 1996, it would be possible to inflate the cost estimates below by 10 to 14 percent to reflect 2000 year equivalents.

It is important that focus be kept on the overall costs of screening and not on single components. For example, it would be a mistake for third party payers to assume that an estimate of laboratory costs to perform the DNA test reported here should be equivalent to any given laboratory's overall charge. The laboratory charge would also include the provision of educational materials, sample collection and/or shipment and, possibly, follow-up counseling. In addition, the laboratory needs to include an additional component to allow for ongoing development. Lastly, a laboratory's advertised ‘list price' is not representative of the actual price charged to clients. Discounted charges are widely available, usually based on volume.

Components 1 and 2: Providing education/information and obtaining informed consent

Component 1: A reasonable cost of providing education/information to pregnant women/couples is about $3.00 (1996 dollars). According to published studies, the median cost of providing such education/information is $2.48 (range $0.80 to $3.33). The costs do not appear to differ significantly based on the screening model chosen. Only one estimate is from the US and it is relatively high ($3.28). Several studies do not include overhead costs. Component 2: A reasonable cost of obtaining informed consent is about $10.00 (1996 dollars). Three reports (all from the US) assume the need for specialized genetic counselors or physicians to provide education/information and to collect informed consent and estimate higher costs. The ACOG/ACMG Guidelines recommend that prenatal cystic fibrosis screening be integrated into routine prenatal care. This means that office staff rather than highly trained counselors will be utilized. Thus, the remaining lower estimates are more likely to be appropriate. Three studies provide a cost for the more explicit dialogue necessary for informed consent. The median value is $4.23 (range $3.79 to $8.60). None of these estimates is from the US, and several do not include overhead costs.

When offering prenatal cystic fibrosis screening to pregnant women or couples, it is essential to offer information about the potential benefits and risks. In pilot trials, this has been accomplished through a combination of pamphlets, videos, and face-to-face communication (Question 38). Several of the economic studies have bundled the provision of education/information with the process of obtaining informed consent. For accurate modeling of costs, it is necessary to separate these two components, since the first is provided to the target population as a whole, while the second is provided only to those who accept screening. The unit cost of providing education/information is multiplied by the number offered testing, independent of how many choose to accept screening. In contrast, the cost of obtaining informed consent is a function of the number who agree to participate.

Most studies assume that information is provided as part of a routine prenatal care visit. This approach has been endorsed by the American College of Obstetricians and Gynecologists (ACOG) and the American College of Medical Genetics (ACMG) (Grody et al., 2001). A few studies assume that information is provided by specialized counselors, which raises costs considerably (Asch et al., 1998; Rowley et al., 1998). Two studies assume the existence of a mass media campaign to publicize and promote prenatal cystic fibrosis screening to the entire population (Ginsberg et al., 1994; Wildhagen et al., 1998), a process which may lower education costs. The remaining studies assume that screening will be offered through practitioners or clinics.

Eight studies have published cost estimates of providing education/information in a prenatal cystic fibrosis screening setting (Table 4-10). Five have bundled together information and informed consent in a single cost. One does not provide a cost for obtaining informed consent. Only one study separately reports the cost of providing information and the cost of obtaining informed consent. The majority of the costs are derived by taking an estimate of staff time and multiplying it by a cost per staff hour. Such estimates may substantially understate costs by failing to include indirect and overhead costs. Pre-screening information provision may differ for couple (or concurrent) screening and sequential screening. For sequential screening, pre-screening information need only be provided to the pregnant woman. For couple (or concurrent) screening, both partners must be informed. Finally, some of the estimates are apparently based on charge rather than actual costs.

Table 4-10. Published Costs of Providing Education/Information and/or Collecting Informed Consent in a Prenatal Cystic Fibrosis Screening Program

Author	Country	Model	Cost	Comment
Combined Education and Consent
Asch	US	Couple	$16.70	Assumes use of genetic counselors
Rowley	US	Sequential	$20.00	Assumes use of specialized counselors
OTA	US	Sequential	$32.23	Assumes use of physician
Providing Education Only
Cuckle	UK	Seq or Couple	Low
Morris	UK	Seq or Couple	$0.86
Ginsberg	Israel	Sequential	$1.24	Assumes mass media campaign
Beech	UK	Sequential	$1.51
Wildhagen	Neth.	Couple	$2.48	Assumes mass media campaign
Beech	UK	Couple	$2.65
Lieu	US	Sequential	$3.28
Obtaining Consent Only
Cuckle	US	Seq or Couple	$3.33
Wildhagen	neth.	Couple	$3.79
Ginsberg	Israel	Sequential	$4.23
Morris	UK	Seq or Couple	$8.60

 

• Rowley and colleagues (1998) state that the unit cost of offering screening is $20 (1996 dollars). This estimate includes professional time (specialized counselors), clerical time, the cost of the informational brochure, and the cost of the consent form. Estimates are derived from a pilot screening program in Rochester, New York. No disaggregated estimates of cost or time are provided.

• Asch and colleagues (1998) assume that the initial offering of screening to a couple will require 37.5 minutes by a genetic counselor. At $26 per hour, this amounts to $16.25 (1995 dollars). This amount does not appear to include the full cost of the genetic counselor's time, including fringe benefits, nor does it include overhead costs, or materials.

• The Office of Technology Assessment (1992) calculates the cost of offering screening to women to be $17 of physician's time and $8 of office staff's time (1991 dollars) and promotional materials. The physician cost is calculated as 10 minutes of time, valued at $100 per hour.

• Ginsberg and colleagues (1994) divide pre-screening information into two components. Publicity for the program is assumed to be a fixed cost of $21,000 per year (1993 dollars). Given 18,522 pregnancies screened and adjusting to 1996 dollars, the cost per pregnancy is $1.24. According to that report, 10 minutes of counseling will be provided before screening is offered. At a labor cost of $10 per hour, this works out to $1.67 in 1993 dollars. No overhead costs are included. No explicit mention is made of obtaining informed consent, but since counseling is provided at a genetics clinic, it is presumably included. Thus, the unit cost of providing counseling prior to testing is $3.87 per couple ($4.23 in 1996 dollars).

• Lieu and colleagues (1994) include a $3 cost (1993 dollars), for pre-test education. This cost is derived from an unspecified project that provided information prior to prenatal screening for Down syndrome or HIV infection. No mention is made of obtaining informed consent from those who choose screening.

• Wildhagen and colleagues (1998) report separate costs for a mass media promotion and individual education of couples choosing screening. The mass media campaign is assumed to cost 136,957 pounds, based on 60 percent of the cost of a media campaign for breast cancer screening. Given 88,241 pregnant couples included, the cost is $2.48 per couple. A second component consists of individual education and is assumed to be twice the observed cost of providing education for breast cancer screening (2.37 pounds, or $3.79 in 1996 dollars). Although not directly mentioned, it is assumed that informed consent was the aim of this second component.

• Beech and Bekker (1995) distinguish the cost of publicizing screening from the cost of encouraging screening. For sequential screening, publicity is accomplished through the distribution of leaflets in prenatal care practices. The cost per pregnancy is estimated to be 0.85 pounds ($1.38 in 1996 dollars). The analysis implicitly assumes a perfect one-to-one distribution of materials to each pregnant woman, with no one missed, no duplicates, and no wastage. For the couple strategy, the cost of publicity rises to 1.49 pounds ($2.41 in 1996 dollars), which covers the cost of postage and mailing of sample collection tubes. The authors assume that with the distribution of educational materials, there is no need to obtain informed consent.

• Cuckle and colleagues (1995) assume that pre-screening education consists of a printed leaflet backed up by an interview with a midwife (Mennie et al., 1992). The cost of creating, printing, and distributing the leaflet was “negligible.” The average time per interview is 10 minutes. It is assumed that a half-time midwife dedicated to this purpose could see 5,000 patients annually, at a cost of 2 pounds per person ($3.33 in 1996 dollars). The study reported that no additional information cost is required for couple screening. Although no mention is made of obtaining informed consent, it is likely that this is the aim of the midwife's interview.

• Morris and Oppenheimer (1995) report separate costs for an informational brochure, at 0.50 pounds ($0.86 in 1996 dollars), and for a 5-10 minute session to obtain informed consent, at 5.00 pounds ($8.60 in 1996 dollars). The study assumes the same costs apply regardless of whether information is being provided to an individual woman or to a couple.

Gaps in knowledge: There are unresolved issues about the cost of pre-screening education and information. The minimum time required to provide detailed information and obtain informed consent is generally regarded to be 10 minutes (Wilfond and Fost, 1992). Key questions include:
1. Is it reasonable to conclude that face-to-face information should only be provided to those who express interest after reading printed materials?
2. Does screening uptake vary with the level of information provided?

Component 3: Obtaining a sample

A reasonable cost for obtaining and shipping a single blood sample is about $10.00. A reasonable cost for obtaining and shipping a single buccal sample is about $4.00. Costs for obtaining buccal samples are lower for two reasons: sampling can be done by the individual, and shipping can be by regular mail. Obtaining a blood sample requires a phlebotomist, and sample shipment is usually by a courier service or express mail.

The cost of collecting blood or buccal samples and transporting them to the laboratory is sometimes not included in analyses of prenatal screening for CF. One study did not include this because almost all blood samples were collected during routine prenatal care for other purposes (Rowley et al., 1998). Some other studies appear to assume that collection cost is included in the laboratory cost of DNA analysis or, it is not considered in the analysis at all. One study includes ten pence as just the cost of providing a tube for the buccal sample (Morris and Oppenheimer, 1995). Another study combines administrative costs with sample acquisition costs (Wildhagen et al., 1998). Table 4-11 shows the four studies that provide a separate cost estimate for sample collection.

Table 4-11. Published Costs of Collecting and Transporting a Sample Specimen as Part of a Prenatal Cystic Fibrosis Screening Program

Author	Country	Sample	Cost	Comment
Asch	US	Blood	$20.00	Charge rather than cost?
Wildhagen	Neth.	Buccal	$14.45	Includes other costs as well
Beech	UK	Blood	$0.62	No shipping costs
Morris	UK	Buccal	$5.37	No shipping costs
		Buccal	$0.44	Includes postage

 

• Asch and colleagues(1998) state that the cost of collecting a single blood sample for DNA analysis is $20, in 1995 dollars. Their analysis includes this as an extra cost in couple-screening strategies but does not apparently include this in their analysis of sequential screening strategies. It is not clear whether this is a cost or charge.

• Wildhagen and colleagues (1998) assume that the costs of collecting and shipping two buccal samples from a couple and administering the screening process amounts to 9.04 pounds per couple, or $14.46 (1996 dollars). No breakdown is provided between collection and administration.

• Beech and Bekker (1996) assume that a nurse working 4 hours per week can collect blood samples from 8,000 patients per year. At an annual labor cost of 19,550 pounds, it costs an average of 0.30 pounds per sample (at a rate of 38 blood samples collected per hour). In addition, they assume a cost of 7 pence for the collection tube. The total cost is $0.62 in 1996 dollars. They do not include shipping costs.

• Morris and Oppenheimer (1995) report that buccal samples can be collected either by post, at a unit cost of 0.30 pounds, or by a cheek swab during a clinical visit, at a unit cost of 5 pound. It is assumed that almost all pregnant women will agree to submit samples by post, for an average cost of 0.31 pounds. The cost of a sample collection tube is put at 0.10 pounds, for a total of 0.41 pounds, or $0.44 (1996 dollars).

Component 4: Performing the DNA test

A reasonable cost for performing the DNA test for about 25 to 30 CFTR mutations in a single blood or buccal sample is $80 to $100. This estimate includes the costs of reagents, supplies, licenses, royalties, technician time, administrative time and overhead. In the future, it is likely that automation and competition will reduce these costs. Although less likely, it is possible that the costs will remain at this level (or decrease less), as technical advances might be used to test for more mutations.

Given the decrease in the costs of DNA analysis resulting from technological progress, adjustments for inflation will not be made for this component. One of the key influences on this cost is the number of mutations included in the panel. According to the ACMG/ACOG guidelines, the panel should include a minimum of 25 mutations. Another key factor is whether estimates are based on actual laboratory costs or the charges from commercial laboratories. Listed charges are considerably higher and include more than the laboratory component. Non-US studies report lower estimates of the cost of DNA testing. The US studies, with one exception, cite commercial charges, rather than actual costs. Most of these studies assume that the costs will be reduced with the introduction of routine screening because of competition, automation and economies of scale. Table 4-12 shows information about DNA testing costs in various publications.

Table 4-12. Published Costs of Performing DNA Analysis as Part of a Prenatal Cystic Fibrosis Screening Program 

Author	Country	Sample	Cost	Comment
Murray	UK		$29
Haddow	US	8-11	$45
Asch	US	6	$50	Based on laboratory costs
Beech	UK	6 or 11	$52
Cuckle	UK	6	$53	Reagents only
Morris	UK	6	$62
Ginsberg	Israel	5	$72	Jewish mutations only
Asch	US	25-35	$100	Projection
Lieu	US	Not stated	$100	Projected cost by experts
OTA	US		$100	Average charge
Rowley	US	70	$150	Commercial laboratory charge
Vintzileos	US		$150	Unknown source

 

• Rowley and colleagues (1998) cite a current charge of $150 for DNA analysis from a single commercial laboratory (Genzyme) testing 70 mutations. They acknowledge that their own laboratory costs are lower.

• Vintzileos and colleagues (1998) assume a DNA analysis cost of $150 for their cost-effectiveness analysis, but no documentation is provided.

• Asch and colleagues (1998) distinguish two testing approaches. A standard analysis includes 5 to 10 common. An expanded analysis includes an additional 20 to 30 mutations. The cost for a six-mutation analysis is reported to be $50 and is based on the costs for technical and professional personnel, reagents, equipment, royalties, laboratory space, and utilities. The expanded battery is assumed to cost $100, but no explanation is provided for this estimate.

• Lieu and colleagues (1994) assume a cost of $100, based on a 1992 survey of experts at four diagnostic laboratories, who projected future charges assuming much higher utilization. The number of mutations tested is not stated.

• The Office of Technology Assessment (1992) conducted a survey in 1991 of laboratories conducting DNA analyses and found an average charge of $170. They use an estimate of $100, including post-test counseling, based on an assumption of economies of scale.

• Haddow and colleagues (1999) cite an unpublished presentation reporting that the cost of DNA laboratory analysis is approximately $45. This estimate does not include laboratory administration.

• Wildhagen and colleagues (1998) use Cuckle's cost estimate of 33 pounds ($53) for multiple mutation analysis. For single mutation analysis of the delF508 mutation, they arbitrarily assume that the cost is one fourth as great, 8.25 pounds, or $13. This is not included in the table because it is derivative.

• Beech and Bekker (1995) report that the cost of a Cellmark Kit used for DNA analysis of CFTR mutations is 30 pounds per person tested. Laboratory staff time to perform each analysis is 2.40 pounds, and laboratory capital cost is 0.11 pounds per person tested, for a total of 32.51 pounds, or $52.02.

• Cuckle and colleagues (1995) report that the cost of the only commercially available reagent kit for testing multiple CFTR mutations is 33.21 pounds. For delF508, an in-house assay in Yorkshire costs 16.10 pounds per specimen, including 6.70 pounds for consumables, 6.20 pounds for staff, and 3.20 pounds for overhead. They use an estimate of 16 pounds for single mutation analyses ($26) and 33 pounds for multiple mutation analyses ($53). The latter estimate does not include any allowance for laboratory staff or overhead. Also, the number of mutations tested is not stated.

• Morris and Oppenheimer (1995) report that the cost of the reagents needed to analyze one sample is 25 pounds. The laboratory and clerical costs add 14 pounds per sample, for a total of 39 pounds, or $62.40. The number of mutations tested is not stated.

• Murray and colleagues (1999) state that a UK commercial supplier of mutation test kits, Zeneca Diagnostics, would provide kits at 12 pounds per test, if purchased in volumes of at least 5,000. Reagents, labor costs, and administrative overhead would raise this to 18 pounds per test, or $28.80 in US dollars.

• Ginsberg and colleagues (1994) report a screening cost of $72.32, based on cost estimates from a single testing laboratory in Israel which includes the costs of labor, raw materials, depreciation and a 16.1 percent overhead.

Component 5: Reporting negative results

A reasonable cost for providing a written report of a negative test result is $2. This covers the cost of clerical time and postage.

Only two studies calculate a cost of reporting test results to everyone screened (Ginsberg et al., 1994; Morris and Oppenheimer, 1995). Other studies implicitly assume that the cost of communicating negative test results is trivial.

Morris and Oppenheimer (1995) report a cost of reporting test results of 1 pound, equally divided between administrative costs and the cost of writing and posting a letter. This cost amounts to $1.73 (1996 dollars).
• Ginsberg and colleagues (1994) report communication of negative test results to an individual woman costs $0.97 in 1993 dollars ($1.06 in 1996 dollars). In addition, 10 percent of those testing negative still request counseling and receive it, at 40 minutes per woman. The cost of staff time is $10 per hour (1993 dollars). The cost per negative test result is therefore $1.64 in 1993 dollars ($1.80 in 1996 dollars).