Characteristics, comorbidities and use of healthcare resources of patients with phenylketonuria: a population-based study.

Abstract Background: Phenylketonuria is a well-known disease, yet the characteristics of the affected population and their use of healthcare resources have not been comprehensively evaluated. Patient characteristics and use of resources are subjects of interest for most governments, especially for a disease included in newborn screening programs. Objective: The aim of this study was to determine characteristics and use of healthcare resources of patients with phenylketonuria in the region of Catalonia. Methods: Records of 289 patients admitted with phenylketonuria between 2007 and 2017 were extracted from the PADRIS database that includes admission data from primary care centers, hospitals (inpatient and outpatient care), extended care facilities, and mental health centers. Results: The patient population was composed of 140 male patients and 149 female patients, and 102 patients were registered via newborn screening during the study period. Patients were admitted on average 2.19 times per year, mostly into primary care centers which concentrated the largest portion of direct medical expenses. Similar percentages of urgent and scheduled admissions were registered both in primary care and hospitals. Annual direct medical cost of treating patients with phenylketonuria was €667 per patient. Finally, 66.80% of the patients suffered from chronic conditions affecting two or more systems, likely to correspond to a wide variety of conditions. Conclusions: Altogether, phenylketonuria patient demographics and direct medical costs in Catalonia have been revised. Patients diagnosed with phenylketonuria appeared 1.3-times more likely to suffer from chronic conditions in distinct organ systems, which is expected to have an effect on their use of healthcare resources. These results support the need to adapt and improve the healthcare system, taking multimorbidity into consideration in an effort to control the medical expenses derived.


Introduction
Phenylketonuria (PKU) is a genetic condition in which the enzyme that processes phenylalanine, phenylalanine hydroxylase (PAH), is deficient or absent. As a consequence, phenylalanine accumulates in the blood, causing a series of symptoms that vary in accordance with age of diagnosis and underlying genetic heterogeneity. Once hyperphenylalaninemia (HPA) is detected, the patient is tested for PAH deficiency and to exclude a non-phenylketonuric HPA caused by the deficiency of its cofactor, tetrahydrobiopterin (BH4) 1 .
When PKU is properly diagnosed the first days of life, patients benefit from nutritional, clinical and biochemical follow-up, and few symptoms are observed; contrarily, when undiagnosed, symptoms include eczema, seizures, autism, behavioral issues, and, eventually, severe intellectual disability 2 . Disease prevalence varies among geographical regions. In Europe, it affects one per 10,000 live births 3 , although in Turkey it increases to one in 4,000 4 . On the contrary, in Finland, prevalence is as low as one per 100,000 live births 5 . In Spain, the geographical distribution of the underlying mutations causing PKU has been analyzed, revealing differences between Mediterranean descendants and originals from the north-west of the country 6 .
PKU is currently diagnosed the first days of life due to the well-established newborn blood screening. Such screenings were established in Spain in 1978 (Real Decreto 2176/1978, August 25) and were soon implemented in the whole country 7 . Their benefits have been evident, with practically all symptoms of the disease being reduced only by dietary restriction 8 . Phenylalanine-restricted diets are still the primary treatment for PKU, often in combination with sapropterin, which reduces blood phenylalanine, or enzyme substitution therapy 9 . In general terms, PKU is a well-known disease, yet the characteristics of the affected population and possible regional discrepancies have not been comprehensively evaluated. Similarly, patients' use of healthcare resources in a certain region is a subject of interest for most governments, especially for the diseases part of newborn screening programs. In such cases, neonatal screenings are justified based on the reduction of costs associated with treating patients later on 10 . A previous study analyzed the costs associated to PKU in Spain, together with its epidemiology 11 ; however, in Spanish regions such as Catalonia (7.5 million inhabitants) the regional government is in charge of the healthcare system, hence the interest of regional evaluations. In order to facilitate the management and examination of patient data, the local government has developed a specialized program that recruits patient information: the data analysis for research and innovation in health program (PADRIS) 12 .
With this in mind this study was designed to evaluate the current status of PKU in the region of Catalonia, focusing on analyzing patient characteristics, revising disease management, and calculating the direct medical costs of the disease.

Materials and methods
Records of all registered patients with a diagnosis of PKU in the region of Catalonia (7.5 million inhabitants) between 2007 and 2017 were analyzed. Data was obtained from project PADRIS via the approval from the University of Barcelona's Bioethics Commission. Parameters such as health centers and medical history identifiers were re-coded prior to extraction to maintain records anonymized in accordance with the principles of Good Clinical Practice and the Declaration of Helsinki.
The database includes detailed information on patients' use of healthcare resources, comprising primary care centers, hospitals (inpatient and outpatient care), emergency rooms (ER), extended care facilities and mental health centers. Records in the database are validated automatically via the evaluation of data consistency. Patient diagnoses and procedures were determined by means of the 9th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-9). When necessary, the extraction of single-patient information was carried out by eliminating repeated records corresponding to separated admissions, relying on the first admission registered during the study period as the index event.
The classification assigns patients an adjusted morbidity group (GMA), according to the number of systems they have affected by chronic diseases, and into five levels of complexity or risk. As it increases, GMA level correlates with a higher number of healthcare admissions of these patients and major pharmaceutical costs 13 . Direct medical costs associated with the use of healthcare resources were calculated based on the mean costs of medical procedures determined by the Catalan government for the year 2013 14,15 .
Data presentation is mainly descriptive. Statistical analyses were performed using Microsoft Excel Professional Plus 2010 (Microsoft Corporation, Redmond, WA).

Results
The study evaluated records of 289 patients diagnosed with PKU in the region of Catalonia between 2007 and 2017 (Table 1). Not significant differences were observed in the percentage of male and female patients, 48.44% vs 51.56%. Surprisingly, patients' mean age at first admission was 14.90 years (SD ¼ 18.07), although age median was 8 years. In total, 102 patients included in the study were registered after diagnosis via newborn screening.
A total number of 15,018 healthcare admissions were analyzed, corresponding to primary health centers, hospitals, ER, extended care facilities, and mental health consultations. Data was gathered between the years 2007 and 2017, averaging 2.19 annual admissions per patient. Significantly, 8.78 admissions were registered annually in primary care centers and only 0.23 annual admissions per patient were into hospitals.
All available information was used to determine the origin of admissions in primary care and hospitals, and the nature of emergency admissions ( Figure 1). A similar proportion of scheduled and urgent visits were observed both in hospitals and primary care, while admissions into ER were mainly due to minor conditions with no vital risk associated.
Mean hospitalization time was 4.57 days, while the mean time patients spent in extended care facilities was 9.8 days. Generally, admissions into ER required stays of less than a day. After hospital discharge, 95.03% of the patients returned to their residence, a percentage slightly lower, 89.73%, in ER admissions ( Table 2).
The Catalan Institute of Health (ICS) assigns patients a value according to the number of systems they have affected  (Figure 2). A 5.13% and 2.66% increase was observed in each of these two categories since 2014. Two distinct approaches have been used to approximate the costs that PKU patients represent for the healthcare system. In terms of disease complexity or risk associated with patients' condition, in 2016 there were seven times more patients in this group at very high risk, and 67% of patients displayed a certain level of risk vs the 50% in the general population ( Figure 3). This data suggests that increased medical costs will derive from the treatment of patients with PKU in comparison with the general population.
On the other hand, an approximation to the direct medical costs of treating patients with PKU was made based on the mean costs of medical procedures established by the regional government (Table 4). Resources destined to primary healthcare were estimated based on the mean cost per admission according to its urgency. Thus, primary attention costs were the most relevant, summing e641,134. ER costs were calculated equally. Hospitalization costs were estimated per admission in outpatient consultations, and based on the average days of stay (1 day or more) and the cost per day. Finally, resources associated to stays in extended care facilities were determined by the cost per admission. Altogether, the annual direct medical cost per patient equaled e667, a total amount that reflects the cost to treat a patient with PKU.
Finally, patients' contribution to pharmaceutical expenses was considered of significance to infer medical costs. Patients are assigned a co-payment percentage according to income, age, and disease. The majority of patients, 54.21%, contributed to 40% of pharmaceutical expenses ( Table 5). 8.30% of patients were exempted from contribution due to age or long-term unemployment.

Discussion
The present study revised all admission records of patients registered with PKU between 2007 and 2017 in the region of Catalonia. Altogether, results showed a predominance of admissions into primary care centers vs secondary care, with an annual cost per patient of e667. A significant frequency of multimorbidities was also measured among these patients.
The analysis revealed few differences to previous evaluations in the whole country. Patients' age at first admission, 14.90 years, was relatively elevated in comparison with the 4.5 years determined in Spain 11 . However, patients' diagnoses are registered in the database solely after 2007, and diagnoses made before that year are not included. Thus, patients born and diagnosed before the study period elevate this parameter preventing a correlation with PKU's first diagnosis age.
Patients were admitted on average 2.19-times per year into a healthcare center, mostly into primary care centers. The Spanish healthcare system, as many others, promotes the use of primary care as a method to increase prevention and maximize healthcare resources 16 . In addition, the characteristics of PKU and its relatively simple control through diet determine a need for general attention received in primary care facilities, and a minority need for urgent attention 2 , consistent with the results of this study.
In terms of secondary diagnoses and comorbidities, no clear tendencies have been identified. A number of patients suffer from chronic conditions such as diabetes type I and II, hypertension, or asthma, but GMA indicators reflect much higher percentages; 66.8% of the patients had chronic conditions affecting two or more systems in 2016, while in the general population this percentage was 51.7% in the same year 17 . The low percentages obtained in the analysis of    secondary diagnoses in admission data suggest that patients' comorbid conditions correspond to a great variety of diseases. This data supports the findings of previous studies that described an elevated presence of multimorbidities in PKU patients affecting various organ systems, including conditions such as allergic and chronic rhinitis, esophageal disorders, overweight, anemia, gastroesophageal reflux disease, dermatitis and eczema, all significantly higher than matched non-PKU controls 18 . Therein, it is speculated that HPA could have consequences beyond neurologic, neurocognitive and neuropsychiatric conditions, but further research is required to evaluate that option. Equally, the restrictive diet necessary to control PKU could have consequences in other organ systems as differences in parameters such as fat oxidation have been confirmed 19 .
On the other hand, the typical neurologic manifestations of an untreated PKU 20 were not detected in this study; the low prevalence of these conditions indicated that the elevated age of first admission does not reflect age of diagnosis in the region.
In general terms, multimorbidity groups have been associated with an increased complexity of disease management and, subsequently, increased medical costs 21 . Additionally, in Catalonia, the ICS has determined a correlation between the presence of multimorbidities and an increased risk and subsequent use of healthcare resources 13 . Health systems are rarely prepared to handle multimorbidity 22 , hence the interest of governments in evaluating disease costs and their relation with multimorbid patients.
The direct medical costs of PKU patients summed e1,348,364 between 2007 and 2017, with a mean annual cost per patient of e667. This cost is presumed higher than in the general population, as a consequence of the major use of healthcare resources that derives from the higher portion of patients with chronic comorbid conditions. In Spain cost estimations situated annual costs of PKU at around e4,239 in 2015, however, therein, calculation methods were based on patients' diagnoses and are not entirely comparable 11 . Currently, the majority of costs associated with the disease are indirect costs, covered by families and patients, due to the extended use of a dietary treatment 23,24 , although prospective treatments facing all disease outcomes might increase medical costs 25 , dramatically rising as well by the presence of comorbid conditions. Finally, a number of limitations could have influenced the results of this study. Data recording in the distinct healthcare sources did not begin uniformly in 2007, and no previous information is available for patients born before that year. Moreover, patients with PKU may not need medical attention in relation with the illness for several years, and a number of patients can remain unregistered in the database.

Conclusions
The characteristics and use of healthcare resources of patients with PKU in the region of Catalonia have been revised. The increased presence of multimorbid conditions affecting distinct organ systems supports the need to evaluate the weight of secondary symptoms derived from this illness. Further research will be required to determine the importance of multimorbid conditions in healthcare use patterns, which will be essential in order to adapt and improve healthcare systems.

Declaration of funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of financial/other relationships
The author declares no competing interests. One reviewer discloses serving as an advisory board member for Biomarin, Applied Pharma Research SA, and Nutricia and receiving speaker fees from Merck Serono, Biomarin, Nutricia, Vitaflo, and Cambrooke. The peer reviewers Abbreviation. ER, Emergency Room.