Peritoneal Dialysis International, Vol. 37, pp. 165–169www.PDIConnect.com

0896-8608/17 $3.00 + .00Copyright © 2017 International Society for Peritoneal Dialysis

165

ECONOMIC IMPACT OF A PERITONEAL DIALYSIS CONTINUOUS QUALITY IMPROVEMENT PROGRAM IN COLOMBIA

Dilip U. Makhija,1 Surrey M. Walton,2 Juan P. Mora,3 and Rafael M. Sanabria4

Baxter Healthcare Corporation,1 Deerfield, IL, USA; Second City Outcomes Research, LLC.,2 Chicago, IL, USA; Coomeva EPS,3 Cali, Colombia; and Scientific Affairs Department,4 Renal Therapy Services,

Latin America, Bogotá DC, Colombia

ORIGINAL ARTICLES

♦ Background: There is little information regarding the financial burden of peritonitis and the economic impact of continuous quality improvement (CQI) programs in peritoneal dialysis (PD) patients. The objectives of this study were to measure the costs of peritonitis, and determine the net savings of a PD CQI program in Colombia. ♦ Methods: The Renal Therapy Services (RTS) network in Colombia, along with Coomeva EPS, provided healthcare resource utilization data for PD patients with and without peritonitis between January 2012 and December 2013. Propensity score matching and regres-sion analysis were performed to estimate the incremental cost of peritonitis. Patient months at risk, episodes of peritonitis pre- and post-CQI, and costs of CQI were obtained. Annual net savings of the CQI program were estimated based on the number of peritonitis events prevented.♦ Results: The incremental cost of a peritonitis episode was $250. In an 8-year period, peritonitis decreased from 1,837 episodes per 38,596 patient-months in 2006 to 841 episodes per 50,910 patient-months in 2014. Overall, the CQI program prevented an estimated 10,409 episodes of peritonitis. The cost of implement-ing the CQI program was $147,000 in the first year and $119,000 annually thereafter. Using a five percent discount rate, the net present value of the program was $1,346,431, with an average annual net savings of $207,027. The return on investment (i.e. total savings—program cost/program cost) of CQI was 169%.♦ Conclusion: Continuous quality improvement initiatives designed to reduce rates of peritonitis have a strong potential to generate cost savings.

Perit Dial Int 2017; 37(2):165–169 epub ahead of print: 28 Sept 2016 https://doi.org/10.3747/pdi.2016.00111

KEY WORDS: Peritonitis, peritoneal dialysis, continuous quality improvement, costs, Colombia.

Peritoneal dialysis (PD) is a safe and effective option for removing waste products in patients without functioning

kidneys. Peritoneal dialysis has also been shown to be less expensive and more cost-effective than hemodialysis (HD), and this may be especially important in lower- and middle-income countries where individual and societal resources are constrained and where access to hemodialysis is limited (1–3). Moreover, compared with HD, patients receiving PD are more likely to rate their dialysis care as excellent (4). As a result, the use of PD in lower- and middle-income countries is gaining in popularity. In 2008, there were approximately 196,000 PD patients worldwide, or about 11% of the dialysis population. Among these, 59% were in developing countries, and the proportion of dialysis patients treated with PD in developing countries increased 2.5-fold between 1997 and 2008 (5). In Colombia, for example, 31.3% of patients on dialysis were reported as being treated with PD in 2012, and over 50% of new starts received PD (6,7).

While PD is less expensive and more accessible than hemo-dialysis (HD), a potential complication of PD is peritonitis (8). Peritonitis can cause significant morbidity, catheter loss, and even death. It has been estimated that peritonitis is the lead-ing cause of PD failure and is responsible for between 1% and 6% of deaths associated with PD worldwide (9). In developed countries, patients may switch to HD if peritonitis occurs, and the rate of transfer from PD to HD is highest in the first 1 or 2 years of PD (10). However, HD is not always an option in lower- and middle-income countries.

Peritonitis in PD typically occurs as a result of touch contam-ination with pathogenic skin bacteria or from catheter-related infection. Patient education and training about optimal tech-niques for PD is one way to help prevent peritonitis. Targeted programs to improve the quality of PD and prevent peritonitis have been effective (11–15). Yu and others reported that a continuous quality improvement (CQI) program improved

Correspondence to: Dilip Makhija, 1 Baxter Pkwy, Deerfield, IL, USA 60015.

dilip_makhija@baxter.comReceived 26 April 2016; accepted 8 July 2016.

by guest on April 8, 2019

http://ww

w.pdiconnect.com

/D

ownloaded from

166

MAKHIJA et al. mARCH 2017 – VOL. 37, NO. 2 PDI

patient survival (16). However, more information is needed to understand whether such programs can consistently reduce peritonitis rates (15). Further, it is not known whether the cost of such programs exceeds the additional costs and resource utilization associated with peritonitis. While CQI programs can address more than just peritonitis, understanding the net benefit of these programs on peritonitis is important to potential policy decisions regarding the implementation of such programs and to broader strategies surrounding expan-sion of the use of PD.

The purpose of this study was 2-fold. First, we sought to determine the cost of peritonitis in patients being treated with PD and, second, to estimate the net economic benefit to the health system of a CQI program designed to prevent peritonitis among patients on PD in Colombia.

METHODS

We conducted a retrospective observational cohort study of PD patients to determine the cost of an episode of peritonitis, and then used those estimates along with cost and impact data from an ongoing PD CQI program to model the net economic benefit of CQI in PD in Colombia. As noted, Colombia is a coun-try with a relatively high PD utilization rate, the reasons for which have been previously described by Sanabria and others (17). In Colombia there is a network of renal treatment facilities called Renal Therapy Services (RTS) that have implemented a PD CQI program. The network includes 42 dialysis centers, 8 PD clinics, and 2 predialysis clinics spread across Colombia and serves 43.5% of the dialysis population, including 4,246 PD patients. The PD CQI program was implemented in RTS clinics in 2007. Six key elements of the program were as follows: 1) an improved ratio of PD nurses to patients, from 1:60 to 1:45; 2) the use of a protocol for exit-site care; 3) standardization of guidelines and protocols for PD dialysis exchanges, patient training, and cleaning of hands and surfaces; 4) certifica-tion and continuing education for PD nurses; 5) in-home support with home visits at 30 days, 90 days, and every 6 months; and 6) systematic follow-up on clinic results with corrective action.

COST OF PERITONITIS

For the period from January 1, 2012, to December 31, 2014, patient-level data were available for patients seen at RTS clinics in Colombia. Only patients 18 years of age or older who were receiving PD were included. Renal Therapy Services clinic records included patient demographic and clinical information, costs of supplies and medications, dates of nephrologist and nursing visits, and events (peritonitis episodes). As the RTS records did not include hospital visits, a subset of patients from RTS that were also in Coomeva EPS were identified and linked to create a de-identified data set for the analysis. The research protocol was reviewed and approved by the ethics boards of RTS and Coomeva EPS in Colombia as well as by the Western institutional review board (IRB), which concurred with ethics boards that patient consent was unnecessary for this project.

We determined the average cost of a peritonitis episode by comparing the costs between those that experienced peritonitis (cases) and those that did not (control). We used propensity score matching to ensure that the groups were equal in terms of other measured factors that may influence costs (see Table 1). To do this we calculated the propensity to develop peritonitis with a logistic regression model based on the available patient characteristics (including age, body mass index [BMI], duration of PD, and comorbid conditions), and then matched cases and controls (in a ratio of 1:1) based on the individual propensity scores using a nearest-neighbor matching strategy with calipers. Once matched, we determined the average weekly cost of care (all visits, hospitalizations, medications, etc.) associated with peritonitis in the cases and the average weekly cost of care in the absence of peritonitis in the controls. The difference was the weekly cost of perito-nitis, which was converted to a per episode cost based on the average duration of peritonitis. The duration of peritonitis was determined based on antibiotic prescription information. All cost estimates were calculated in Colombian Pesos (COP) and then converted to US dollars using an exchange rate of 0.00033 available at the time of the analysis (18). The above analyses were conducted using Stata, version 13 (StataCorp LP, College Station, TX, USA).

TABLE 1 Characteristics of Groups Before and After Matching

Before matching After matching Cases Controls P Cases Controls P Characteristic (n=201) (n=860) value (n=199) (n=199) value

Male, n (%) 116 (57.7) 487 (56.6) 0.78 115 (57.8) 105 (52.8) 0.31Age (yrs), mean (SD) 56.5 (14.3) 56.7 (16.1) 0.87 56.4 (14.3) 53.5 (17.8) 0.07BMI, mean (SD) 25.7 (5.1) 25.0 (4.2) 0.04 25.7 (5.1) 25.5 (4.3) 0.72Duration of PD (yrs), mean (SD) 1.7 (1.6) 1.4 (1.6) 0.04 1.7 (1.6) 1.7 (1.9) 0.95CAPD, n (%) 110 (54.7) 478 (55.6) 0.83 110 (55.3) 120 (60.3) 0.31Diabetes, n (%) 102 (50.8) 418 (48.6) 0.59 100 (50.3) 87 (43.7) 0.19Hypertension 71 (35.3) 316 (36.7) 0.71 70 (35.2) 56 (28.1) 0.13

SD = standard deviation; BMI = body mass index; PD = peritoneal dialysis; CAPD = continuous ambulatory PD.

by guest on April 8, 2019

http://ww

w.pdiconnect.com

/D

ownloaded from

167

PDI mARCH 2017 – VOL. 37, NO. 2 EVALUATION OF A CQI PROGRAM IN PD

IMPACT OF PROGRAM

As mentioned above, the PD CQI program was implemented in 2007 in all the RTS clinics in Colombia, and aggregated data on the program were collected starting the year prior (2006) and for the 8 years after implementation (2007 – 2014). These data included the number of patient months (volume of ser-vice), the number of episodes of peritonitis, and the cost to provide the program based on internal calculations by RTS of the investment made to improve the nurse-patient ratio, edu-cational activities provided, protocol and guideline settings, educational tools for patients, and support materials such as fixation belts for home therapy.

As the primary purpose of the PD CQI program was to pre-vent peritonitis, any cost savings incurred by the program was considered to result from reductions in healthcare expenses (including inpatient and outpatient care) that would have been required for treatment of peritonitis. Specifically, to determine the cost saving of the PD CQI program in each year, we multiplied the average cost per episode of peritonitis (as described above) by the estimated number of cases of peritonitis averted because of the program in that year. The cost of running the program was then subtracted from this to get the net economic benefit each year.

The number of peritonitis episodes averted each year was determined by comparing the actual number of perito-nitis episodes with the expected number of episodes. The expected number of episodes was estimated by multiply-ing the rate (episodes per patient-month) in 2006 (before implementation of PD CQI program) by the number of patient-months in each subsequent year (2007 – 2014). As the rate of peritonitis could have decreased over time due to other changes in healthcare or the environment, we also calculated a more conservative estimate of episodes averted that assumed a 5% annual reduction in the rate of peritonitis.

To determine the overall net economic benefit of the program we summed the discounted present value of the net savings each year from 2007 – 2014. The equation used for net benefit was

NB = Σyear (t1- t8) [(E t × CE t) – CPt],

where NB is net economic benefit ($), E = peritonitis episodes averted in that year, CE = the cost per peritonitis episode prevented, and CP = the cost of the program in that year. We also calculated the percent return on investment (ROI) of the CQI program for the 2007 – 2014 period. The equation used for ROI was

ROI = Σyear (t1- t8) [((E t × CE t) – CPt) / CPt] × 100.

The discount rate used was 3%. The data analysis and calculations for this section of the research were done using Microsoft Excel, 2010, version 14.0.7165.5000 (Microsoft Corporation, Redmond, WA, USA).

RESULTS

COST OF PERITONITIS

There were a total of 1,061 patients who met the inclusion criteria, among whom 201 experienced at least 1 episode of peritonitis during the follow-up period. There were 199 cases that were matched based on propensity score, with 199 controls (non-cases). The characteristics of cases and controls before and after matching are shown in Table 1. Before matching, the peritonitis group had a slightly higher BMI (25.7 vs 25.0, p = 0.04) and a longer duration of PD prior to the study period (1.7 years vs 1.4 years, p = 0.04), but these differences were resolved after matching. There were no other significant dif-ferences in observable characteristics prior to matching.

In looking at average costs overall during an episode, including usual care costs, nurse visits comprised 11% of the episode costs, nephrologist visits 21% of the episode costs, and hospital-based costs comprised the remainder. The largest components of the hospital costs were medications (34%), hospital materials and supplies (20%), and PD solutions (19%). Of those who experienced an episode of peritonitis during the follow-up period, the average duration was 22.2 days (± 11.7) with an interquartile range of 15 – 29 days. During a peritonitis episode, patients had an average of 4.0 (± 3.6) nurse visits and an average of 2.7 (± 2.2) nephrologist visits. In comparing average costs per week for patients during an episode with those without an episode in the matched cohorts, the additional cost per week for an episode of peritonitis was $78.87 (95% confidence interval [CI] 70.87 – 86.81), which corresponded to a per episode cost of $250.13.

IMPACT OF THE PROGRAM

The cost of implementing the CQI program in Colombia for approximately 4,000 patients was $147,000 (year 1), and main-taining the program cost $119,000 per year thereafter. In spite of a relatively low cost, we observed a sharp decline in perito-nitis rates across time after implementation. The frequency of peritonitis in 2006 was roughly 1 case per 21 patient-months, and this improved to 1 case in 60 patient-months by 2014. Relative to levels seen in the year prior to implementation, over 8 years, the program prevented 10,409 episodes of peritonitis (mean 1,301 per year). Figure 1 illustrates the decline in peri-tonitis with CQI relative to a patient-month-adjusted baseline used to approximate peritonitis in the absence of CQI, as well as a scenario where peritonitis would have declined by 5% per year without CQI. In the more conservative scenario, there were 7,052 episodes of peritonitis prevented by the program over 8 years (mean 881 per year).

Considering the estimated episodes of peritonitis prevented and the cost per episode of peritonitis, and subtracting the cost of the program, the discounted net economic value of the PD CQI program over 8 years was $1,346,431 with an average (undiscounted) annual savings of $207,027. This is a ROI of 169%. Under the more conservative scenario where we

by guest on April 8, 2019

http://ww

w.pdiconnect.com

/D

ownloaded from

168

MAKHIJA et al. mARCH 2017 – VOL. 37, NO. 2 PDI

assumed that other improvements in PD equipment or tech-nique would reduce peritonitis by 5% per year independent of the CQI program, the discounted net economic value of the program over 8 years was $679,143 (average annual savings $100,734). The ROI of the program under this scenario was approximately 82%.

DISCUSSION

Peritoneal dialysis remains an important treatment alterna-tive for patients requiring dialysis, especially in areas where access to hemodialysis is limited by travel barriers or unavail-ability. However, a limitation of the use of PD is the risk of peritonitis. Beyond the morbidity and mortality associated with peritonitis, it is associated with additional healthcare costs. To the best of our knowledge this is the first study to examine both the additional cost of peritonitis and the net economic benefit that a CQI program can have by reducing peritonitis episodes.

Using data from PD patients seen in renal clinics in Colombia, South America, we found that the expected additional cost of each episode of peritonitis was $250. Very little has been published on the cost of peritonitis. In fact, we could find only 1 other study, based on older data. Piraino and others reported a cost of peritonitis resulting from catheter infection ranging from $152 to $5,122 based on data from 1986 to 1987 (19).

The PD CQI program in Colombia that we describe appears to have been able to reduce peritonitis episodes substantially. Other investigators have reported similar findings. For example Yu and others reported that a PD CQI program improved the rate of peritonitis from 1 episode in 22.2 patient-months

to 1 episode in 30.1 patient-months in the first year after implementation and, subsequently, to 1 episode in 71.1 and 77.3 patient-months in years 2 and 3, respectively (16). Wang and others reported a reduction in the incidence of peritonitis from 1 in every 40.1 patient-months before the CQI to once every 70.8 patient-months after CQI (15).

The PD CQI program described in this study did not differ substantially in its components from other PD CQI programs that have been described in the literature (11,14–16). While these types of programs exist elsewhere, they have not pre-viously been examined for economic impact. Although our data are specific to Colombia, given the relatively low cost of implementation of PD CQI relative to the burden generated by peritonitis, it is likely that PD CQI would create net benefits to other systems. Further, our findings were robust across a scenario where alternative environmental factors could have reduced peritonitis by 5% per year independently from the PD CQI program.

Our data were limited to patients in Colombia within a specific health maintenance organization (HMO) provider, and the data for determining the cost of peritonitis were limited to a short time period (2012 – 2013). Hence the estimates may not generalize to other settings or other countries, and if the cost of peritonitis has varied across time, the actual net benefit in earlier years may be over- or underestimated by using the cost of peritonitis that we found in 2013. Further, the estimates of net benefit for the CQI program relied on a historical control, and there were no data available from those sites to test prior trends or trends in sites without CQI. Other unobservable factors that may have changed across time and that were correlated with peritonitis could have biased the results. However, our results were robust to an alternative scenario that assumes a 5% per year decline in peritonitis would have occurred without the program. Further, the fact that the cost of the CQI program was low relative to the cost of peritonitis is likely to be true in other settings. Hence, there is strong reason to believe that impactful programs can generate substantial savings within a health system.

CONCLUSION

Peritonitis is a potential adverse event of PD associated with added cost. However, evidence suggests that impactful CQI programs can generate a substantial net benefit for health systems. Moreover, in systems that have implemented CQI to reduce peritonitis, there is further incentive for the use of PD relative to HD.

ACKNOWLEDGMENTS

This study was sponsored by Baxter Healthcare Corporation. The study design and data analyses were supported by Baxter Healthcare Corporation through contracts with Second City Outcomes Research LLC, a consulting company in which Surrey M. Walton is a managing partner. Much of the data came from Renal Therapy Service (RTS), a subsidiary company of Baxter Healthcare Corporation. Rafael M.

Figure 1 — Figure 1 shows the impact of a PD CQI program on episodes of peritonitis after implementation. The blue line represents the number of episodes of peritonitis expected in the absence of the PD CQI program. This was determined based on the rate of peritonitis in the year before implementation. The red line represents the expected number of episodes of peritonitis in the absence of the PD CQI pro-gram, with the additional assumption that the rate would be reduced by 5% annually from improvements in healthcare unrelated to the PD CQI program. The green line is the actual number of episodes of peritonitis that occurred. The difference between the green and blue line, or the green and red line, can be attributed to the PD CQI pro-gram. PD = peritoneal dialysis; CQI = continuous quality improvement.

by guest on April 8, 2019

http://ww

w.pdiconnect.com

/D

ownloaded from

169

PDI mARCH 2017 – VOL. 37, NO. 2 EVALUATION OF A CQI PROGRAM IN PD

Sanabria participated in the collection of data, some of the plan-ning around data analysis, and assisted in the interpretation of the results. Further, Baxter approved the decision to submit the article for publication. In addition, we wish to acknowledge assistance with analyses and writing from Glen Schumock and Todd Lee of Second City Outcomes Research LLC.

DISCLOSURES

We have read and understood Peritoneal Dialysis International’s policy on disclosing conflicts of interest and declare the following interests: Dilip U. Makhija is an employee and shareholder of Baxter Healthcare Corporation. Rafael M. Sanabria is an employee of Renal Therapy Services, a subsidiary company of Baxter Healthcare Corporation. Surrey M. Walton has received fees for consulting from Baxter via Second City Outcomes Research LLC. This work was presented as an abstract and poster at the 2016 Congress of the International Society for Peritoneal Dialysis (ISPD), Melbourne, Australia. REFERENCES

1. Neil N, Walker DR, Sesso R, Blackburn JC, Tschosik EA, Sciaraffia V, et al. Gaining efficiencies: resources and demand for dialysis around the globe. Value Health 2009; 12:73–9.

2. Just PM, de Charro FT, Tschosik EA, Noe LL, Bhattacharyya SK, Riella MC. Reimbursement and economic factors influencing dialysis modality choice around the world. Nephrol Dial Transplant 2008; 23:2365–73.

3. Treharne C, Liu FX, Arici M, Crowe L, Farooqui U. Peritoneal dialysis and in-centre haemodialysis: a cost-utility analysis from a UK payer perspec-tive. Appl Health Econ Health Policy 2014; 12:409–20.

4. Pacheco A, Saffie A, Torres R, Tortella C, Llanos C, Vargas D, et al. Cost/utility study of peritoneal dialysis and hemodialysis in Chile. Perit Dial Int 2007; 27:359–63.

5. Jain AK, Blake P, Cordy P, Garg AX. Global trends in rates of peritoneal dialysis. J Am Soc Nephrol 2012; 23:533–44.

6. Sanabria M, Muñoz J, Trillos C, Hernández G, Latorre C, Díaz CS, et al. Dialysis outcomes in Colombia (DOC) study: a comparison of patient survival on peritoneal dialysis vs hemodialysis in Colombia. Kidney Int Suppl 2008; (108):S165–72.

7. System USRD. USRDS 2012 annual data report: Atlas of chronic kidney disease and end-stage renal disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2012.

8. Voinescu CG, Khanna R. Peritonitis in peritoneal dialysis. Int J Artif Organs 2002; 25:249–60.

9. Fried LF, Bernardini J, Johnston JR, Piraino B. Peritonitis influences mor-tality in peritoneal dialysis patients. J Am Soc Nephrol 1996; 7:2176–82.

10. Chaudhary K. Peritoneal dialysis drop-out: causes and prevention strate-gies. Int J Nephrol 2011; 2011:434608.

11. Li J, Wang H, Xie H, Mei G, Cai W, Ye J, et al. Effects of post-discharge nurse-led telephone supportive care for patients with chronic kidney disease undergoing peritoneal dialysis in China: a randomized controlled trial. Perit Dial Int 2014; 34:278–88.

12. Shetty A. Reducing peritoneal dialysis-related peritonitis rate. Ochsner J 2014; 14:386–91.

13. Yao Q, Zhou G. Role of a center of excellence program in improving the quality of peritoneal dialysis—a Chinese experience. Perit Dial Int 2014; 34(Suppl 2):S59–62.

14. Fang W, Ni Z, Qian J. Key factors for a high-quality peritoneal dialysis program—the role of the PD team and continuous quality improvement. Perit Dial Int 2014; 34(Suppl 2):S35–42.

15. Wang J, Zhang H, Liu J, Zhang K, Yi B, Liu Y, et al. Implementation of a continuous quality improvement program reduces the occurrence of peritonitis in PD. Ren Fail 2014; 36:1029–32.

16. Yu Y, Zhou Y, Wang H, Zhou T, Li Q, Li T, et al. Impact of continuous quality improvement initiatives on clinical outcomes in peritoneal dialysis. Perit Dial Int 2014; 34(Suppl 2):S43–8.

17. Sanabria M, Devia M, Hernández G, Astudillo K, Trillos C, Uribe M, et al. Outcomes of a peritoneal dialysis program in remote communities within Colombia. Perit Dial Int 2015; 35:52–61.

18. Anon. The money converter. Vol. 2015.19. Piraino B, Bernardini J, Johnston J. Cost analysis of peritoneal catheter

infections. Perit Dial Int 1990; 10:241–2.

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.For commercial re-use, please contact marketing@multi-med.com

by guest on April 8, 2019

http://ww

w.pdiconnect.com

/D

ownloaded from