Quality assurance, efficiency indicators and cost-utility of the evaluation workup for liver transplantation

Authors


Abstract

We report the results of a retrospective review of the outpatient pretransplantation workup for United Network for Organ Sharing (UNOS) 3 patients adopted at a liver transplantation (LT) center and illustrate the efficiency indicators used for quality evaluation and cost-analysis. A single-center, pre-LT evaluation workup was performed on an outpatient basis at a cost per patient evaluation of 2,770 Euros (€). Objective measures were: the number of patients admitted to and excluded from each phase of the algorithm; the rate of patients admitted to pre-LT evaluation out of the total of referred patients (the referral efficiency rate); the rate of waitlisted patients out of those admitted to pre-LT evaluation (the evaluation efficiency rate); the rate of waitlisted patients out of those referred for LT (the process efficiency rate); and the cost per waitlisted patient, as the ratio of the cost per patient evaluation to the evaluation efficiency rate. From January 1, 1996, to October 1, 2004, 1,837 patients were referred for LT on an outpatient basis. Based on preemptive evaluation of the available clinical data, 412 patients (22.4%) were excluded from pre-LT evaluation and 1,425 (77.6%) were admitted to preliminary consultation. Among these, 603 (42.3%) were excluded from and 822 (57.7%) were admitted to pre-LT evaluation with a referral efficiency rate of 44.7% (822 of 1,837). Out of the patients evaluated for LT, 484 were waitlisted with a cost-utility and evaluation efficiency rate of 58.8% each (484 of 822). Of the 1,837 patients originally addressed for LT 484 were waitlisted, yielding a process efficiency rate of 26.3% (484 of 1,837) and a cost per waitlisted patient of €4,710.8. In conclusion, the 3 indicators allowed monitoring of the efficiency of the pre-LT evaluation algorithm. The current process efficiency rate at our center is low (26.3%), but avoiding early referrals we might increase it to 31.6%, with a 12% net saving on costs per waitlisted patient (from €4,710.8 to €4,165.4). (Liver Transpl 2005;11:1080–1085.)

In recent years, liver transplantation (LT) has become established in the management of acute and chronic liver diseases and has been the object of extensive cost analysis evaluation. However, much attention has been paid to cost determinants of cadaveric and living donor LT procedures for acute and chronic liver failure in the adult or pediatric setting,1–13 to cost-effectiveness of cytomegalovirus infection prophylaxis,14–17 and treatment strategies for posttransplantation hepatitis B virus recurrence.18–22 Only scanty data are available on cost-effectiveness of retransplantation23; treatment of posttransplantation hepatitis C virus recurrence24; screening, monitoring and transplantation for hepatocellular carcinoma (HCC)25–27; techniques of biliary anastomosis28; LT in obese patients29; sclerosing cholangitis30; management of acute graft rejection31; immunosuppressive strategies32, 33; and posttransplantation intensive care unit stay.34, 35

The pretransplantation workup plays a crucial role in the whole LT algorithm, and efficiency evaluation is mandatory to improve standards of care and conform health care delivery strategies to patients' needs and societal requirements. However, quality assurance requires identification of efficiency parameters to allow assessment of compliance of the pretransplantation workup with the process endpoints (i.e., LT), single out criticalities, and plan consequent improvements. The aim of the current study is to report the results of a retrospective review of the pre-LT algorithm adopted at our center for United Network for Organ Sharing (UNOS) 3 patients, illustrate the process efficiency indicators used for quality evaluation, and highlight strategies to improve the process cost-utility. To the best of our knowledge this is the first description of pretransplantation workup efficiency parameters in the international literature.

Abbreviations

LT, liver transplantation; HCC, hepatocellular carcinoma; UNOS, United Network for Organ Sharing.

Materials and Methods

The Liver Transplant Program of Pisa University started on January 1, 1996. In October 2004, a retrospective review of the outpatient clinic activity of the Liver Transplant Unit was carried out. The review consisted of chart evaluation of all UNOS 3 patients referred between January 1, 1996, and October 1, 2004, and whose clinical data were stored in a prospectively maintained nurse data base (Excel for Windows, Microsoft, Seattle, WA). Apart from urgent patients and those referred for inpatient care (UNOS status 1, 2a, and 2b), the pre-LT evaluation workup was performed on an outpatient basis from 4 to 5 days at a cost per patient evaluation of 2,770 Euros (€). This cost includes all laboratory tests, radiologic investigations, and medical consultations contemplated by the evaluation protocol in use at our center, but does not include the preliminary visit with the transplant surgeon or hepatologist. For the purposes of the current study the outpatient clinic algorithm was subdivided into 4 sequential steps (Fig. 1): patients' referral and preemptive evaluation of the available clinical data; preliminary patients' consultation; pre-LT workup; and process endpoint (LT waitlisting). Objective measures were: the number of patients admitted to and excluded from each phase of the algorithm; the rate of patients admitted to pre-LT evaluation out of the total of referred patients (the referral efficiency rate); the rate of waitlisted patients out of those admitted to pre-LT evaluation (the evaluation efficiency rate); the rate of waitlisted patients out of those referred for LT (the process efficiency rate); and the cost per waitlisted patient, as the ratio of the cost per patient evaluation to the evaluation efficiency rate (Table 1). The primary study endpoints were evaluation of process efficiency and cost-analysis through adoption of the aforementioned indicators. The secondary study endpoints were to single out system criticalities in the outpatient clinic algorithm and quantify the impact of appropriate organizational strategies on the efficiency and cost-utility of the pre-LT evaluation process. Data are reported as means, medians, or percentages, as necessary. The costs of the pre-LT evaluation workup were calculated on the basis of the 2004 edition of the Nomenclatore Tariffario Regionale (the Regional Health Care Reimbursement Catalog) disciplining reimbursement for private and public health care delivery practice in Tuscany, and updating fees according to yearly inflation rates. According to the Italian National Institute of Statistics (ISTAT) the mean yearly inflation rate in the period from 1996 to 2003 was 2.4% ± 1.4% (median 2.45%; range 1.6%-3.9%).36

Figure 1.

Flowchart of the pre-LT evaluation algorithm for UNOS 3 patients of the Liver Transplant Unit of Pisa University (1996 to 2004).

Table 1. Efficiency and Cost-Utility Indicators of the Pre-LT Evaluation Algorithm Adopted at the Liver Transplant Unit of Pisa University
IndicatorDefinition
Referral efficiency ratePatients admitted to pre-LT evaluation out of the total of referred patients
Evaluation efficiency ratePatients waitlisted out of the total of patients admitted to pre-LT evaluation
Process efficiency ratePatients waitlisted out of the total of referred patients
Cost per patient evaluationCost of the pre-LT evaluation work-up
Cost per waitlisted patientCost of the pre-LT evaluation work-up per waitlisted patient, as the cost per patient evaluation/evaluation efficiency rate

Before the advent of the Model for End-Stage Liver Disease, eligibility to LT and patient prioritization were based on the UNOS scoring system and time of entry into the waitlist, respectively. Since 2002, patient prioritization has been based on both UNOS and Model for End-Stage Liver Disease scores. LT for HCC is indicated within the Milan criteria,37 while patients outside these criteria are referred to LT on an individual basis, including age, hepatitis C virus status, HCC stage, histology (when available), and response to pre-LT transarterial chemoembolization. Concurrent extrahepatic malignancies of any origin are considered contraindication to LT, apart from in situ neoplasms. A past medical history of low-grade malignancies is not considered contraindication to LT when the relapse-free survival is ≥3 years. Abstinence from alcohol ≥6 months is required for eligibility to LT. Obesity with a body mass index ≥28 kg/m2 is a contraindication to LT at our center, while age per se is not considered a major contraindication to LT unless it is associated with concurrent morbidities.

Results

From January 1, 1996, to October 1, 2004, a total of 548 LT were performed on 512 patients. In the same period 3,018 patients were referred to our outpatient clinic; among these were 1,837 (60.9%) UNOS 3 patients for evaluation of eligibility to LT. Figure 1 illustrates the number of patients admitted to and excluded from each phase of the algorithm from referral to LT waitlisting.

Based on preemptive evaluation of the available faxed and/or transmitted clinical data, 412 patients (22.4%) among the ones referred were excluded from subsequent consultation with the transplant physician because of existing contraindications to LT in 256 cases (62.1%) and early referral in 156 (37.9%). Contraindications to outpatient consultation were: HCC beyond the center criteria in 29.7% of cases (76 of 256); major cardiovascular and/or pulmonary disease in 25.7% (66 of 256); active alcohol abuse and/or non compliance with alcohol abstinence ≥6 months in 14.1% (36 of 256); concurrent and/or past extrahepatic malignancies according to the center criteria in 7.4% (19 of 256); secondary hepatic malignancies other than neuroendocrine tumors in 6.6% (17 of 256); cerebrovascular and/or neurologic disease in 6.6% (17 of 256); obesity in 4.7% (12 of 256); active drug abuse in 3.1% (8 of 256); decompensated insulin-dependent diabetes mellitus with target organ failure in 1.5% (4 of 256); and hepatitis C virus-related end-stage renal failure in 0.4% (1 of 256). Early referrals consisted of a Child-Turcotte-Pugh score ≤7.

Of the 1,837 patients originally referred, 1,425 (77.6%) were admitted to the outpatient clinic for preliminary evaluation of eligibility to LT for a mean of 13.6 ± 3.4 patients per month (median 14; range 5-24). After consultation 603 patients (32.9%) were excluded from subsequent pre-LT workup: 356 cases (59%) for contraindications to LT and 247 (41%) because of early referral. The contraindications to pretransplantation evaluation were: HCC beyond the center criteria in 34.5% of cases (123 of 356); major cardiovascular and/or pulmonary disease in 23.3% (83 of 356); active alcohol abuse and/or noncompliance with alcohol abstinence ≥6 months in 12.6% (45 of 356); secondary hepatic malignancies other than neuroendocrine tumors in 10.6% (38 of 356); systemic infection of nonhepatic origin in 7.3% (26 of 356); obesity in 4.5% (16 of 356); concurrent and/or past extrahepatic malignancies according to the center criteria in 3.6% (13 of 356); active drug abuse in 1.9% (7 of 356); and systemic vasculitis in 1.4% (5 of 356). Early referrals consisted of a Child-Turcotte-Pugh score ≤7.

Of the 1,425 patients admitted to the outpatient clinic, 822 (57.7%) underwent the pre-LT evaluation workup in the same period for a mean of 7.8 ± 1.3 patients per month (median 7.5; range 2-11). Among the 822 patients undergoing the pre-LT workup, 484 (58.8%) were waitlisted and 338 (41.1%) were excluded from waitlisting, in 244 cases (72.2%) for comorbidities, and in 94 (27.8%) were for early referral. The contraindications to LT waitlisting: major cardiovascular and/or pulmonary disease in 34.8% of cases (85 of 244); HCC beyond the center criteria in 30.3% (74 of 244); cerebrovascular and/or neurologic disease in 10.6% (26 of 244); active alcohol abuse in 6.5% (16 of 244); active infection of non-hepatic origin in 6.1% (15 of 244); concurrent extrahepatic malignancies according to the center criteria in 2.4% (6 of 244); psychological non compliance in 2% (5 of 244); active drug abuse in 1.6% (4 of 244); secondary hepatic malignancies other than neuroendocrine tumors in 1.2% (3 of 244); hepatitis C virus–related chronic renal failure in 0.4% (1 of 244); and patient refusal to be waitlisted in 3.6% (9 of 244). Early referrals were due to the absence of previously reported HCC in 79.8% of cases (75 of 94) and a Child-Turcotte-Pugh score ≤7 in 20.2% (19 of 94).

Out of the 1,837 patients addressed to our outpatient clinic for evaluation to LT, 822 (44.7%) underwent the pre-LT workup and 484 (26.3%) were waitlisted (Fig. 1). Of the 1,353 patients (73.6%) excluded from the pre-LT evaluation algorithm at any time point, 856 cases (63.3%) were for existing contraindications to LT and 497 (36.7%) for early referral. Thus, the referral efficiency rate was 44.7% (822 of 1,837), the evaluation efficiency rate 58.8% (484 of 822), and the process efficiency rate 26.3% (484 of 1,837) (Table 2). Based on the 2004 edition of the Nomenclatore Regionale Tariffario overall reimbursement fees for all pre-LT workups ranged as high as €2,276,940 (€2,770 × 822), i.e., €21,606 ± €3,601 per month, while the cost per waitlisted patient—defined as the ratio of the cost per patient evaluation to the evaluation efficiency rate—was €4,710.8 (€2,770/0.58) (Table 2). Ideally, if all early referrals (497 of 1,837) had been appropriately identified prior to pre-LT evaluation and not addressed to our center (Fig. 2), the referral efficiency rate would have been 54.3% (728 of 1,340), the evaluation efficiency rate 66.5% (484 of 728), and the process efficiency rate 36.1% (484 of 1,340). Under such circumstances total costs of pre-LT workups would have been €2,016,560 (€2,770× 728) and the cost per waitlisted patient €4,165.4, with a net saving of €565.4 for each waitlisted patient.

Table 2. Results of the Efficiency and Cost-Utility Indicators of the Pre-LT Evaluation Algorithm Adopted at the Liver Transplant Unit of Pisa University
IndicatorRatioResult
Referral efficiency rate822/183744.7%
Evaluation efficiency rate484/82258.8%
Process efficiency rate484/183726.3%
Cost per patient evaluation€2,770
Cost per waitlisted patient€2,770/0.58€4,710.8
Figure 2.

Based on our experience, the figure illustrates an ideal patient flowchart if all early referrals had been appropriately identified before being addressed to our center.

Discussion

Despite its relevance to the whole LT algorithm, little attention has been paid to the pretransplantation patient workup and to the necessary health care strategies to improve its quality and cost-utility. However, timely and efficient evaluation of transplant candidates is crucial to improve the efficacy of LT and conform it to patients' needs. No universal agreement exists on the best referral patterns for UNOS 3 patients eligible for LT, and no cost-analysis has been reported to this respect in the international literature. Moreover, quality assurance of the pre-LT workup requires identification of reliable parameters to appraise its efficiency within and across LT centers and monitor the impact of health care strategies. The aim of our retrospective study was not only to review the outpatient referral algorithm for UNOS 3 candidates in use at our center, but also to illustrate the efficiency indicators we adopted for quality evaluation and suggest their application in current LT practice.

The process efficiency of our outpatient evaluation algorithm—expressed as the number of patients waitlisted for LT out of the total of referred patients—was low (26.3%), but no data are available for comparison in the international literature. Intuitively, this low yield was mainly due to futile and/or early referrals (referral efficiency rate 44.7%), since the percentage of patients excluded from the pretransplantation workup after preemptive data evaluation and/or preliminary consultation (55.2%, i.e., 1,015 of 1,837) greatly exceeded that of patients who were not waitlisted after the pretransplantation workup (18.4%, i.e., 338 of 1,837). Therefore, reducing the percentage of futile and/or early referrals might have a more significant impact on the efficiency of the whole algorithm than further refinement of the pretransplantation workup. Given the extensive pretransplantation workup in use at most LT centers, increase of the evaluation efficiency rate might be possible through modification of patient referral patterns, rather than adding up investigations and/or laboratory tests to the evaluation algorithm.

Since eligibility for LT may vary across centers and greatly affect the percentage of futile referrals, strategies to increase both the referral efficiency and the process efficiency rates may consist of a working partnership between the LT center and the referring institutions and/or physicians with diffusion of the center eligibility criteria throughout the LT center catchment area. A further strategy to improve the process efficiency might be avoiding patients seeking consultation with transplant physicians without previous medical advice. Efficient transmission of all available clinical data and subsequent preemptive patient chart evaluation by transplant physicians might reduce futile indications and/or early referrals. Our experience shows that only 22.4% of patients (412 of 1,837) were excluded from the LT evaluation workup upon preemptive chart evaluation, but such a percentage might greatly increase if clinical data were made available to the LT center for all referred patients prior to consultation.

The ideal algorithm should be the one with a referral efficiency rate of almost 100%, where all referred patients apparently match the LT center eligibility criteria and are admitted to the pretransplantation workup. This result is difficult to achieve in current LT practice, since it would require preemptive identification of early referrals and contraindications to LT. However, by avoiding that early referrals are addressed to our center, we might increase the evaluation efficiency rate from 58.8% to 66.5% and the process efficiency rate from to 26.3% to 36.1%, with a 12% saving on costs per each waitlisted patient (from €4,710.8 to €4,165.4), and a net saving of €273,653.6 for the whole waitlisted population (€565.4 × 484 waitlisted patients), i.e., €4,410.1 ± €735.02 per month. Therefore, the economic impact of an efficient cooperation between the LT center and the referring physicians is invaluable in the current era of budget constraints, and every effort should be undertaken to make the pre-LT evaluation process as efficient and cost-useful as possible.

In conclusion, at our center introduction of efficiency and cost-utility indicators—such as the referral efficiency rate, the evaluation efficiency rate, the process efficiency rate, the cost per patient evaluation, and the cost per waitlisted patient—allowed us to single out process criticalities and plan appropriate strategies to improve the pre-LT evaluation algorithm for UNOS 3 patients. Currently, no parameters exist to appraise the efficiency of the pre-LT evaluation process and we suggest the application of the reported indicators for within- and across-center comparison. Areas of further investigation should be cost-effectiveness of the pretransplantation evaluation algorithm, its impact on LT outcome, and patients' survival rates.

Acknowledgements

The authors thank Monica Masetti, Cinzia Tellini, Rosa Paganelli, and Juri Ducci for assistance in retrieving data.

Ancillary