Impact of a Lung Transplantation Donor-Management Protocol on Lung Donation and Recipient Outcomes

Lung transplantation is often the only available treatment option for patients with end-stage lung disease. Because of important advances in surgical techniques and pharmacologic management, lung transplantation offers many patients an improved quality of life and an increased likelihood of survival (1). Despite these advances, however, few patients benefit from transplantation because of the scarcity of lung donors.

In 2005, nearly 3,000 patients were on the national waiting list for lung transplantation. Only 35% of these patients received transplants, and 10% died while awaiting a graft (2). Approximately half of the listed patients waited more than 2 yr before receiving a transplant. This trend worsens each year; the number of new additions to the waiting list annually is nearly double the number of patients who receive transplants (2).

The lack of organ donors is most serious for patients awaiting lung transplantation, in part because lungs are procured from only 16% of organ donors (2). Multiple factors influence this low procurement rate; among them are acute lung injury after brain death, pneumonia, aspiration, and atelectasis (3, 4). In addition, many potentially usable organs are excluded because of current standard criteria for lung donors that have been in existence for more than two decades (4-8). Several strategies have been proposed for reducing the shortage of lung donors, including increasing communication between transplant centers and organ procurement organizations (OPOs), performing active medical management of potential lung donors, and modifying (or extending) the currently accepted criteria for donor selection (6, 7, 9-15). These published strategies contributed to our development of the San Antonio Lung Transplant (SALT) donor-management protocol. We hypothesized that implementing the SALT protocol would increase the lung procurement rate without adversely affecting the overall survival rate of lung transplant recipients. The principal aim of this study was to determine the impact of the SALT donor-management protocol on the rates of lung donation and the clinical outcomes of lung transplant recipients. Some of the results of this study have been previously presented in abstract form (16-19).

METHODS

Methods, Patients, and Study Sites

In September 2001, we implemented the SALT donor-management protocol for all potential organ donors evaluated for candidacy by the Texas Organ Sharing Alliance (TOSA) in San Antonio, Texas. The TOSA medical board and the Institutional Review Board of the University of Texas Health Science Center at San Antonio approved the research protocol. TOSA provided data regarding the procurement of donor organs for the period from September 1, 1997, through August 31, 2005; thus, we were able to compare data from the 4 yr before initiation of the SALT protocol with data from the 4 yr after its initiation. Donors older than 65 or younger than 10 yr of age and non-heart-beating donors were excluded from this review. The remaining donors were considered potential lung donors. Two of the investigators (J.E.C., G.B.C.), who were not involved with the care of any of the donors or recipients, applied our revised donor criteria and recipient outcomes to the appropriate databases.

SALT Protocol Components and Implementation

Education. Our transplant pulmonologists (L.F.A., D.J.L.) met with the TOSA staff for training sessions on donor selection and management. The main objectives of the training sessions were (1) to emphasize that every organ donor should be regarded as a potential lung donor and that requesting and obtaining consent for lung donation should be attempted for every organ donor and (2) to provide education about donor-management strategies.

Active donor evaluation and management. After consent had been obtained for organ donation, all potential lung donors from any of the referring hospitals were managed by our transplant pulmonologists and the TOSA staff. Management strategies included performing ventilator recruitment maneuvers, restricting fluid administration, administering diuretics, and implementing techniques for preventing aspiration (20). After consent for lung donation had been obtained, we implemented active alveolar recruitment when initial blood gas analyses demonstrated that the Pa^sub O2^/Fl^sub O2^ ratio was less than 300, when pulmonary infiltrates were consistent with pulmonary edema or atelectasis, or when both conditions were present (21). Alveolar recruitment involved pressure-controlled ventilation at an inspiratory pressure of 25 cm H2O and positive end-expiratory pressure of 15 cm H2O for 2 h. We then switched the ventilator mode to conventional volume control ventilation with a tidal volume of 10 ml/kg and a positive end-expiratory pressure of 5 cm H2O (22, 23). Recruitment was deemed successful if Pa^sub O2^/Fl^sub O2^ improved to greater than 300 and the chest radiograph demonstrated substantial improvement after 30 min of conventional ventilation (final Pa^sub O2^/Fl^sub O2^). No other potential recruitment measures were used during the study periods. We clinically assessed the donor fluid balance, minimized the use of crystalloids, and recommended the administration of diuretics to maintain a neutral or negative fluid balance after the initial hospital resuscitation. To decrease the risk of aspiration, we elevated the head of the bed to 30

Modified donor selection criteria. Lung donor criteria were categorized as absolute or extended. The following were absolute donor criteria: Pa^sub O2^/Fl^sub O2^ > 300, no radiographically visible infiltrates, and no evidence of copious purulent secretions or bronchoscopically demonstrated aspiration. Extended donor criteria included one or more of the following characteristics: age more than 55 yr, cumulative smoking history of more than 20 pack-years, history of pulmonary disease, severe chest trauma, more than 4 d on mechanical ventilation, and positive results from Gram staining of tracheal or bronchoalveolar lavage fluids. These extended criteria included fixed characteristics that could not be altered by medical management.

Revised donor classification. Potential lung donors were classified as "ideal donors" if they met all of the absolute criteria and demonstrated none of the extended criteria, as "extended donors" if they met all of the absolute criteria but demonstrated one or more of the extended criteria, and as "poor donors" if they did not satisfy all of the absolute criteria. After implementation of the SALT protocol, potential donors who were initially classified as poor donors could remain so classified (and thus no longer be considered for lung donation) if the protocol's donor-management strategies did not improve their status; alternatively, they could be considered "poor to ideal donors" or "poor to extended donors" if the donor-management strategies improved their status from poor to "ideal" or "extended."

Donor selection for lung transplantation. Once the donors had been classified, the TOSA coordinators performed a match run and called the transplant institutions to offer one or two lungs for transplantation. The recipient institutions had access to the donor records and a verbal report from the coordinator about the donor. They were informed of the donor-management and recruitment maneuvers performed on each donor and the results of those maneuvers. The decision to accept any of the offered organs from ideal, extended, poor to ideal, or poor to extended donors was made by the lung transplant surgeons from the recipient's institution; each institution had its own set of criteria that could vary from that the criteria at other institutions. The type of transplantation procedure (single-lung transplantation versus bilateral transplantation) was determined by the recipient's characteristics and diagnoses, not by the donor type. Neither the TOSA personnel nor the pulmonologists associated with the protocol were involved in any recipient institution's decision to accept lungs for transplantation.

Data Abstraction

We retrospectively collected data for the pre-SALT period (September 1, 1997, to August 31,2001) and prospectively collected data for the SALT period (September 1, 2001, to August 31, 2005). We used the TOSA database and donor charts to obtain donor data, such as demographic information, comorbid conditions, intensive care unit (ICU) records, cause of admission and death, routine laboratory values (complete blood count, electrolytes, liver function, and blood gases), and intubation details (field or hospital, evidence of aspiration during intubation). We collected recipient data from all patients who received transplants at our institution until the time of their death or for a minimum of 1 yr after the procedure. We collected the following data from lung transplant recipients: demographic information, indication for transplantation, comorbid conditions, ICU data, routine laboratory values, pulmonary function tests, and survival rates.

Endpoints

For the donor groups, our primary outcome was the procurement rate of lungs for transplantation from potential organ donors. The secondary outcome was the consent rate for lung donation. For the recipients of lung transplants from each of the four types of donors (ideal, extended, poor to ideal, and poor to extended), the primary outcome was survival at 30 d and at 1 yr. secondary outcomes included mean Pa^sub O2^/Fl^sub O2^ as measured 24 h after the transplantation procedure, the number of mechanical ventilation days, length of stay (LOS) in the ICU and in the hospital, and results of pulmonary function tests 1 yr after the transplantation procedure.

Statistical Analyses

The demographic characteristics of donors in the pre-SALT and SALT periods and the clinical and laboratory findings were analyzed with simple t tests and ?2 tests. Confidence intervals (CIs) for simpler proportions and means were calculated by using the Wald method. Exact Poisson methods were used to compute the means and 95% CIs for lung procurement rates. An interrupted time-series analysis was used to evaluate the effects of the SALT protocol in this situation because randomization was not feasible. This type of analysis is particularly suited to evaluate the impact of changes in health care policy or to evaluate the impact of interventions designed to change medical practices (24, 25). The objective was to establish whether the change observed after implementation of the SALT protocol produced a stable change in the procurement rate, as opposed to natural cyclical variation over time. We fitted an interrupted time-series model to the number of donors procured each month via a generalized estimating equation model with number of donors as a Poisson random variable (26). A Poisson error model with a log-link function was assumed, and the scale parameter was computed as the deviance divided by the residual degrees of freedom for the purpose of correcting for overdispersion or underdispersion in the distribution. The fitted model was Y = ß^sub 0^ + ß^sub 1^Month + ß^sub 2^(Month - 48)(Intervention) + ß^sub 3^intervention. Months were labeled from 0 to 96; the donor-management protocol was implemented in Month 48. The parameter ß^sub 3^ estimates the difference in mean response during the pre-SALT period and the SALT period, whereas ß^sub 2^ tests the difference in slopes between the pre-SALT trend and the SALT trend. Exposure was defined as the number of potential donors evaluated for candidacy during each 30-d period.

Kaplan-Maier survival estimates were determined for lung transplant recipients during the pre-SALT and SALT periods. The recipients were classified by donor type: ideal, extended, poor to ideal, and poor to extended. Differences between the groups were assessed with a logrank test; statistical significance was assigned at p < 0.05. A bootstrap multiple comparison procedure with Bonferroni adjustment was used to compare recipients for differences in intubation days, LOS in the ICU, total LOS in the hospital, 24-h Pa^sub O2^/Fl^sub O2^, and FEV^sub 1^ at 12 mo. Analyses were performed with SAS statistical software (SAS Institute, Inc., Cary, NC), Stata (StataCorp LP, College Station, TX), and R (Free Software Foundation, Inc., Boston, MA).

RESULTS

We implemented the SALT protocol in September 2001. A total of 711 potential lung donors were evaluated for candidacy at 38 different medical centers during the 8-yr study period: 330 before the introduction of the SALT protocol (pre-SALT) and 381 after its initiation (SALT). Sixty-four percent of the potential lung donors were evaluated for candidacy at six large medical centers. The groups were similar in most baseline characteristics: The mean donor age was 36 yr, and the cause of brain death was equally divided between traumatic and medical conditions (Table 1). Approximately 60% of the donors underwent intubation at the hospital; the mean length of intubation before pronouncement of brain death was 2.5 d. The mean Pa^sub O2^/FI^sub O2^ before brain death was significantly lower (p < 0.0001) for the SALT donor group than for the pre-SALT donor group (Table 1).

Endpoints: Lung Donation and Transplantation Rates

The mean lung procurement rate from all potential donors was 11.5% (95% CI, 8.1-15.0) during the pre-SALT period and 25.5% (95% CI, 21.1-29.8) during the SALT period; there was an estimated risk ratio of 2.2 (95% CI, 1.1-4.4; p < 0.001) in favor of the SALT period (Figures 1 and 2). The number of potential lung donors who were consented for donation was significantly higher during the SALT period (87%) than during the pre-SALT period (76%; p < 0.001), as was the number of consented donors who became actual lung donors (15.2% pre-SALT; 29.2% SALT; p < 0.0001; Figure 1). The rate of lung offers from potential donors to transplant centers was 21% during the pre-SALT period and 57% during the SALT period (p < 0.001). Despite this increase in the rate of lung offers, there was no significant difference in the number of lungs rejected by transplant centers (43% pre-SALT; 55% SALT; p = 0.10). The procurement rate of lungs per lung donor was similar during both periods (1.57 lungs per donor pre-SALT; 1.7 lungs per donor SALT; p = 0.75). As a result of the increased number of donor lungs, the number of patients who received a lung transplant increased significantly, from 53 during the pre-SALT period to 121 during the SALT period (p < 0.0001; Figure 1). Thirty-eight percent of the lung transplantations were performed at eight transplant centers outside the TOSA region during the pre-SALT period, whereas 33% were performed at 14 transplant centers outside the region during the SALT period (p = 0.60).

Donor Classification

After implementation of the SALT protocol, 254 (76%) lung donors were initially classified as poor, 37 (11%) as ideal, and 42 (13%) as extended donors (Table 2). The most frequently observed characteristic of the donors initially classified as poor donors was Pa^sub O2^/FI^sub O2^ < 300 (67% of donors), followed by infiltrates on chest radiograph (58%) and abnormal results from bronchoscopy (28%). Sixteen poor donors had only abnormal results from bronchoscopy, with Pa^sub O2^/FI^sub O2^ > 300 and no findings demonstrated by chest radiography.

Among the 124 extended donors, the most commonly seen extended characteristic was a positive result from Gram staining (80 of 124; 65%), followed by smoking status (23 of 124; 19%) and mechanical ventilation for more than 4 d (22 of 124; 18%). Of the 98 actual lung donors, 35 (36%) demonstrated only one extended characteristic, 13 (13%) demonstrated two extended characteristics, and 4 (4%) demonstrated more than two extended characteristics.

Impact of the SALT Protocol on Poor Donors

The group of donors initially categorized as poor composed the largest subgroup from which lungs were procured during the SALT period. Of the poor donors, 135 (53%) were reclassified as poor to ideal or poor to extended donors after the implementation of donor-management strategies. The level of hypoxemia and the findings from bronchoscopy were the main variables associated with an increased conversion of poor donors to poor to ideal or poor to extended donors (Table 3). Only 10% (11 of 115) of the poor donors in the pre-SALT period became actual lung donors, whereas 21% (53 of 254) of the poor donors in the SALT period became actual lung donors (p < 0.0001). Of the 121 lung transplantation procedures performed during the SALT period, 64 (53%) used grafts from donors reclassified as poor to ideal or poor to extended, whereas only 14 (26%) lung transplants during the pre-SALT period came from such donors (p = 0.006). A final Pa^sub O2^/FI^sub O2^ > 400, improvement in the Pa^sub O2^/FI^sub O2^ ratio of > 150, and donor age < 45 yr were the main characteristics associated with conversion of poor to ideal or poor to extended donors to actual lung donors (Table 4).

Impact of Donor Management on Oxygenation

We found that the mean Pa^sub O2^/FI^sub O2^ of 59% of all potential lung donors improved during the SALT period (Pa^sub O2^/FI^sub O2^, 240-372; mean change, 131), whereas that of only 31% of the potential lung donors improved during the pre-SALT period (Pa^sub O2^/FI^sub O2^, 263-351; mean change, 87; p < 0.0001). The mean final Pa^sub O2^/FI^sub O2^ of the actual lung donors was significantly better during the SALT period (Pa^sub O2^/FI^sub O2^, 463) than during the pre-SALT period (Pa^sub O2^/FI^sub O2^, 416; p = 0.02). There were no significant differences in the final Pa^sub O2^/FI^sub O2^ between the different donor groups (ideal, extended, poor to ideal, and poor to extended) during the SALT period (see Table 2).

The impact of the SALT protocol on oxygenation was most substantial among the 105 poor donors with Pa^sub O2^/FI^sub O2^ < 300 or with pulmonary infiltrates who underwent recruitment maneuvers (Pa^sub O2^/FI^sub O2^, 236-320; mean change, 90). This change in oxygenation with recruitment maneuvers accounted for 31% (30 of 98) of the actual lung donors.

Recipient Outcomes

During the study period, 114 lung transplantation procedures were performed at our institution, with a minimum follow-up of 1 yr. There was no significant difference in the underlying diagnosis of the recipients. The rate of bilateral lung transplantation increased from 19% during the pre-SALT period to 37% during the SALT period (p = 0.02). The 30-d survival rate (81% pre-SALT; 99% SALT; p = 0.005) and the 1-yr survival rate (76% pre-SALT; 85% SALT; p = 0.14) improved after implementation of the SALT protocol. No significant differences were found during the SALT period in 30-d and 1-yr survival rates between recipients of lungs from the donor groups (ideal, extended, poor to ideal, and poor to extended; Figure 3). There were no significant differences between the recipients of lungs from the four donor groups in the clinical outcomes of oxygenation, days on mechanical ventilation, LOS in the ICU or in the hospital, or FEV^sub 1^ at 1 yr (Table 5).

DISCUSSION

This study addresses one of the primary limitations of lung transplantation: the disparity between the number of actual donors and the number of patients awaiting lung transplantation. Our implementation of the SALT donor-management protocol was associated with significantly larger numbers of donors consented, lungs procured, and lungs transplanted. The key components of the SALT protocol were (1) improving communication between our transplant center and our local OPO, (2) providing direct involvement and active management of donors by our lung transplant physicians, and (3) establishing a donor classification system after management of potential donors. Furthermore, implementing the SALT protocol did not compromise the outcomes of the lung transplant recipients, as evidenced by comparable 30-d and 1-yr survival rates, oxygenation assessments, duration of mechanical ventilation, LOS in the ICU and in the hospital, and lune function.

Improvements in lung procurement with donor-management strategies have also been reported by Gabbay and colleagues from Australia (27). These authors reported that the use of extended-donor organs and the institution of aggressive donor management increased procurement rates from 33 to 51%. A similar experience was reported by the California Organ Bank, where the use of a donor-management protocol has increased procurement rates by as much as 35% over the last few years (2, 28). In those studies, as in ours, the increase in organ donation was not associated with a change in the survival rates of recipients.

A key strategy for increasing the number of useable donor lungs in our study was modifying donor criteria by reassessing oxygenation, chest radiography, and bronchoscopic findings as absolute criteria after active donor management. This study was not focused on the use of extended donors as much as on the successful optimization of potential lung donors.

The use of extended donors has been retrospectively evaluated at other transplant centers, although these centers did not implement any specific protocol for managing potential donors. In 1993, Kron reported the first series of lung transplantation procedures using extended-donor lungs; there were no differences in the immediate postoperative results of 11 lung transplant recipients (29). Some large transplant centers have reported comparable short-term outcomes with similarly liberalized donor criteria (6, 7, 11, 12, 30-32). However, other centers have observed that the use of extended donors resulted in increased 30-d mortality rates among recipients of lungs with bilateral infiltrates, frankly purulent secretions, or both (7); the recipients also required prolonged stays in the ICU and the hospital (33). In contrast, our study found that neither the LOS in the ICU and in the hospital nor the 30-d and 1-yr mortality rates differed between recipients of lungs from the four donor groups.

Good oxygenation among potential lung donors (final Pa^sub O2^/FI^sub O2^ > 300) before donation is a required criterion in most transplant programs (3, 4, 20). During the SALT period, we were able to improve the oxygenation of potential lung donors by implementing the SALT donor-management protocol. In contrast, during the pre-SALT period, when no donor-management protocol was used, oxygenation decreased substantially. The significance of improving oxygenation is highlighted by the fact that in our study a final Pa^sub O2^/FI^sub O2^ > 400 was the strongest donor variable associated with acceptance of lungs by transplant centers. Recruitment maneuvers were an important component of the SALT protocol; when these maneuvers were used, the condition of 21% of the donors who were initially classified as poor donors was improved sufficiently to allow them to become actual lung donors.

This study has some limitations relating to preintervention and postintervention design. We used a retrospective comparison cohort to analyze the impact of the implementation of the SALT protocol. Additionally, organ bank personnel were not required to follow the SALT protocol; thus, compliance with all aspects of the protocol did not always occur. However, although this lack of compliance is considered a study limitation, it reflects the realities associated with any clinical change and increases the likelihood that similar protocols introduced at other local OPOs could have similar results. Finally, as is true of many other studies of lung transplantation, this study involved a relatively small recipient population, and we were unable to collect complete data from recipients undergoing transplantation at other institutions. However, with the observed survival rates among our SALT cohort, achieving the statistical power to detect a 10% difference in mortality rates between the recipient groups would have required the enrollment of nearly 300 patients in each group.

In conclusion, our results demonstrate that implementing the SALT protocol was associated with a significant increase in the number of lung donors and in the number of lung transplantation procedures performed without increasing LOS in the hospital or the ICU and without compromising the pulmonary function or survival rates of the lung transplant recipients. A similar donor-management program, implemented on a large scale by organ banks around the country with a procurement rate approaching 25%, would nearly double the number of lung transplantation procedures, significantly decrease lengthy wait times, and potentially eliminate many of the deaths that occur among patients awaiting lung transplantation.

Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the manuscript.

Acknowledgment: The authors thank Flo Witte, M.A., E.L.S., for editorial assistance.

© 2006 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.