Real-world experience of hereditary angioedema (HAE) in Mexico: A mixed-methods approach to describe epidemiology, diagnosis, and treatment patterns

Introduction and objectives Due to the lack of structured and systematic information available, the aim of this study was to describe the epidemiology, diagnosis, healthcare processes, and treatment patterns of hereditary angioedema (HAE) in Mexico. To achieve this, different data sources were consulted regarding medical literature, structured health system databases, and angioedema-specialized physicians (AEP) opinion regarding HAE. Material and methods A mixed methods approach was conducted in 4 phases: I) systematic literature review (SLR) and meta-analysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines; II) review of national health system (NHS) databases and systematic reports; III) physician survey; and IV) an epidemiologic model. ICD 10 D84.1 encoded records from the NHS were used to estimate the number of patients with HAE attended and treated during 2019. A survey was implemented to increase understanding of the clinical profile and treatments used. Results A prevalence rate of 0.9/50 000 inhabitants was estimated for 2019. In the same year, an estimated 317 HAE type 1 patients were recorded in the NHS, aged ≥11 years old. The most frequent clinical symptoms were cutaneous edema (67.5%) and abdominal pain (47.9%). A severe episode with laryngeal edema appeared in 27.5% of cases. Acute episodes were mainly moderate to severe (77.0%), with an annual per capita frequency of emergency visits of 7.6 patient-year (range 1–12/patient-year). The main reasons for hospitalization corresponded to laryngeal facial, tongue, and abdominal edemas, representing 73.3% of annual ICD 10 D84.1 reported hospitalizations. The main treatments that patients with HAE received were fresh frozen plasma for acute attacks and danazol for short-term prophylaxis (STP). Conclusions Despite efforts to make HAE visible, according to this study, cases recognized and treated in the NHS represent only 16.6% of the estimated prevalence.


INTRODUCTION
Hereditary Angioedema (HAE) is characterized by recurrent, localized, and self-limited angioedema episodes (not accompanied by wheals) involving the skin and/or mucous membranes of the respiratory and gastrointestinal tracts. 1 Recurrent angioedema can be divided into inherited forms, HAE, and acquired forms, acquired angioedema (AAE). 1 A genetic C1 inhibitor (C1-INH) deficiency (HAE type 1) or lack of function (HAE type 2) are the causes of the classic and most common types. 1 HAE is a rare, unpredictable, serious, and potentially fatal genetic disease; 75% are inherited in an autosomal dominant manner, and up to 25% of cases correspond to spontaneous genetic changes (de novo mutation). 2The genetic background of HAE lies fundamentally in SERPING1 gene variants, which generates decrease (HAE type 1) or dysfunction without decrease (HAE type 2) of C1-INH.In HAE C1-INH, around 85% cases are HAE type 1 and 15% cases are HAE type 2. 3 HAE is recognized as a rare disease with a worldwide estimated prevalence in the range between 1 case per 50 000 and 1 case per 100 000 inhabitants, 4 with no identified differences in prevalence due to gender. 5Recently published information points to differences in prevalence among ethnic groups (Hispanics, Asians, white and black populations) 6 and regions (western countries vs Asia Pacific). 7Due to its low prevalence and the fact that it shares symptoms with other more frequent diseases, HAE is not usually suspected in the first instance in some areas such as Mexico and other developing countries.This situation leads to misdiagnosis and underdiagnosis contrasting with Europe and developed nations. 8nsequently, patients are often treated ineffectively and late, living many years (on average 22), with a significant impact on their quality of life (QoL) 9 until finally obtaining a correct diagnosis and appropriate treatment.Some of them have even undergone unnecessary surgeries (for example, negative appendectomy) due to lack of disease awareness. 10other potential reason for the underdiagnosis of HAE is the low availability of population-based epidemiological studies worldwide, as well as the existence of few systematized patient registries* which explains why the prevalence figures that circulate in literature show great dispersion, 9 thus making it difficult to accurately estimate the number of patients with HAE.
[14] According to the National Center for Technological Excellence in Health, and the Mexican Association of HAE (consulted in June 2022), 15 in Mexico there are no population-based epidemiological studies, nor an HAE registry of patients.
Finally, given that the current and available data sources of information in Mexico 16 do not include diagnostic-therapeutic details of HAE patients' care, it is likely that the current estimates of the burden of disease are inaccurate.Therefore, it is also important to compare the shared participation of public and private care for patients with HAE, given the differences in type of patients, budget constraints, and access to both up to date diagnostic methods and specialized care.The aim of this study was to estimate the prevalence in HAE types 1, 2, and nC1-INH, understand the process of diagnosis, medical care, health resources and specific treatments consumption, during acute attacks and prophylactic strategies in Mexico.

METHODOLOGY
A mixed methods approach was conducted in 4 phases: I) Systematic Literature Review (SLR) and meta-analysis; II) review of the national health system (NHS) databases and systematic reports; III) angioedema-specialized physicians (AEP) survey; and IV) an epidemiologic model.

Systematic literature review and meta-analysis
A systematic literature review (SLR) was carried out to obtain information on HAE: prevalence and frequency of HAE types 1, 2, and nC1-INH.This SLR was conducted according to PRISMA guidelines. 17he sources of information used included the following platforms and search engines: electronic library of the National Autonomous University of Mexico (UNAM), National Library of Medicine (NLM, Pubmed), OVID, SCIENCE DIRECT, SciELO and Virtual Health Library (BVS, for its acronym in Spanish).Access to the following data bases was obtained: MEDLINE, EMBASE, LILACS, EMB Reviews, Practice Guidelines, The Cochrane Library, and IMBIOMED.
Literature search was conducted accordingly to find papers with the following inclusion criteria as it is recommended by Moncada-Hernández: 18 Mexican and international peer reviewed journals indicated in the medical databases; HAE prevalence (number patients/50 000-100 000 inhabitants); patients with an established HAE diagnosis by gender (male, female); patients' age (0-100 years old); patients by age group (10, 11-20, 21-30, 31-40, 41-50, 51-60, and >60 years); HAE type (Type 1, Type 2, or nC1-INH); study classification (systematic reviews, case series, descriptive studies of patient populations, and population-based epidemiological studies).Exclusion criteria: all the HAE papers that do not meet the inclusion criteria (not include HAE by type, gender, age groups, or population explicitly) and conference abstracts.
The key search terms used were all crossed referenced with the terms: HAE or Hereditary Angioedema.
The search strategy included studies published in English or Spanish from January 1999 to April 2022.The articles which met the inclusion criteria were searched for additional references.Exact search algorithms with keywords, Medical Subject Headings terms, entree terms and search strings, and details regarding the SLR are available in Supplement S1.Full-text available studies were included for initial review.Articles were reviewed by 2 authors independently, with discrepancies resolved after joint article review and discussion. 18e selected published articles and specialty theses (see details Supplement S2) underwent a meta-analysis with the Comprehensive Meta Ana-lysisÒ program from the US National Institutes of Health (NIH).
From this data, the prevalence rate of HAE/50 000-100 000, and the frequencies of type 1, type 2, type nC1-INH, female, male, and age groups were obtained, and point values and 95% CI (Confidence Intervals) were calculated.

Review of Mexican NHS databases and systematic reports
The objective of this phase was to retrieve from the databases and systematic reports of the NHS (public and private) through the records encoded by ICD 10, the number of HAE cases attended during 2019, the distribution of cases (by public and private healthcare, gender, and age group), as well as the number of outpatient consultations, medical emergency consultations services, and hospitalization data events (number, rate, and days of stay >24h).
The following NHS databases and systematic reports for 2019 were reviewed: the statistical yearbooks of the major health institutions in the public sector (97% coverage), [19][20][21][22] as well as the database of health information in private health units, 23 the annual report of morbidity 24 and mortality, 25 and the report of hospital discharges by specific cause of the general direction of epidemiology of the Ministry of Health (MoH) in Mexico. 26or the tracking of cases treated during 2019 ICD code D84.1 was used, which only identifies cases with deficiency C1-INH.Also, it does not differentiate between inherited and acquired cases.The differentiation between hereditary and acquired cases, as well as type 1 or type 2 HAE was estimated with the clinical information obtained through a survey applied to AEP.
Other public database used correspond to the 2020 Population and Housing Census from the National Institute of Geography and Statistics (INEGI). 27From here, data on the structure of the Mexican population and the population pyramid for each age group and gender were obtained (the adjusted data for 2019, was also retrieved from INEGI). 28 accordance with the above, the estimation of patients cared for in the NHS during 2019 was calculated by the retrieval of records coded with ICD 10 D84.1 from all NHS institutions and the annual report of the private sector for 2019.From the records, the distribution of cases by age group and gender was obtained, as well as the number of outpatient consultations, medical emergency consultations services, and hospitalized patients (number, rate, and days of stay).All the records were added, including the demo-epidemiological and medical care items above mentioned.Thus, the total number of cases registered, attended, and treated in the NHS was obtained.From the total, the share of both the private and public sectors was determined.In the public sector, the share of each one of the institutions was identified.

Physician survey
To understand the clinical status and treatment of patients with HAE, a survey was designed to collect the data from the AEP (Supplement S3).To develop questions and consolidate the survey, a Delphi Panel 29 was implemented.For controlling the quality of Delphi methodology, according to Nasa et al, 29 we committed attention to 9 points: 1) identification of a problem, 2) area of research, 3) panel selection and anonymity, 4) controlled feedback, 5) iterative Delphi rounds, 6) consensus criteria, 7) analysis of consensus, 8) closing criteria, and 9) stability of the results.
We used a convenient sampling strategy to identify AEP to whom the survey was applied.[21][22][23] The AEP were selected from the hospitals corresponding to the major institutions of the public sector, and for the private sector, tertiary care hospitals in Mexico City.The proportion of AEP was determined according to the share of records obtained from the public and private sectors, respectively.[21][22][23] The objective of the survey administered to AEP was to obtain the average clinical profile (signs, symptoms, frequency, and severity of acute attacks), the average treatment used for the acute attack and for short-term prophylaxis (STP) management (aim: prevention of acute attack related to a potential trigger condition, ie, infection, minor or major surgery, etc), as well as validation of the frequency of HAE types, gender distribution, and affected age groups.Also, the proportion of cases of HAE and AAE was estimated with the clinical information obtained through the survey applied to 40 AEP; the relevant data that obtained consensus during the Delphi Panel to filter clinically were: absence of urticaria, recurrent abdominal pain or swelling, lack of response to conventional antiallergic treatment (typically use of histamine Type-2 Receptor Antagonists and glucocorticoids), family history, recurrent angioedema attacks, and average age of symptoms onset (childhood, youth, adult).Other mentioned data ACEI treatment.
Finally, and from the survey also, the differentiation between Types 1 and 2 was made based on laboratory studies carried out (C1-INH levels, C1-INH functional quantification, C1q and C4 levels).The verification rate with the genetic test in relation to tests performed for clinical and laboratory suspicion was also considered.The mean, standard deviation, and 95% CI were calculated from the 40 surveys retrieved from AEP.The reported results correspond to the analysis of the survey responses.

Epidemiologic model
As mentioned above, there are no reliable epidemiological studies nor population registries of patients with HAE for Mexico. 15In this regard, a probabilistic model was built in Microsoft ExcelÒ to obtain specific Mexican calculations.
To estimate the HAE prevalence, a meta-analysis of the information obtained from population studies that considered specific countries and other determined groups was carried out. 30The information was only available in aggregate form at the level of total populations by country or specific groups and documented casesregistered by country or population in the same period.The number cases x 100 000 inhabitants were determined for each country or population analyzed.
According to the frequency determined by the panel of AEP, the population estimate for HAE type 1 and type 2 was made.Since the ratio for HAE nC1-INH agreed upon by the panel of AEP was small, it was decided to add it on top of that determined for HAE types 1 and 2. Mean values and 95% CI were calculated.
The model was populated with the specific information obtained (assumed as proxy variables) from the following information sources: -data from the official Mexican population (2019) The schematic mechanics of the modeling process is shown in Supplement S4, in brief: to estimate prevalence for Mexico, a meta-analysis obtained from population studies that considered Fig. 1 PRISMA diagram of the systematic literature review (SLR) specific countries and other determined groups was carried out.The mean value and scenario with a triangular distribution was determined from the extreme maximum-minimum values available for the populations.With these, health modeling was carried out using a first-order Monte Carlo simulation with 10 000 iterations.Mean values and 95% confidence interval (CI) were estimated as rates.From this, we calculated the number of total HAE cases (C1-INH deficiency) for the total Mexican population in 2019.
Regarding HAE cases, reported in the NHS (determined from the review of the NHS databases and systematic reports, specifically through ICD 10 D84.1 coded records [January 2019-December 2019]), 376 were identified as C1-INH deficiency (public and private) cases.From this and based on what was obtained from the survey applied to AEP, it was estimated that 352 patients were HAE (93.6% [95% CI 91.2%-95.2%]),and 24 (6.4% [95% CI 5.9%-9.3%])cases AAE.From the 352 HAE cases, 317 were estimated as HAE Type 1 (90.1% [95% CI 84.3%-91.1%]),and 35 cases were considered HAE Type 2 (9.9% [95% CI 9.2%-10.4%]).On the other hand, patients with nC1-INH were not considered for the present analysis since there is no specific or suggestive ICD-10 code for them.The AAE cases were not considered for further analysis either.
From the NHS databases and systematic reports consulted relative to the medical care site of these patients, 77.6% of the cases (273) were seen in the public sector: 43.8% (120) in the IMSS (  (National MoH), 10.8% (29) in the ISSSTE (Institute of Security and Social Services for Federal Government Workers) and 3.5% (10) in other institutions.In the private sector 79 patients were cared for 22.3% (Table 2).
The main causes of hospitalization corresponded to facial, tongue, larynx, face, and neck edema (68.8%), and severe edema in the abdomen simulating acute abdomen (8.9%), as well as edema in 2 or more body segments (20%), and important distal limb edema (2.2%).Regarding hospitalization within the NHS, 81.4% of the cases (125) were hospitalized in the public sector, while 18.6% (29) were admitted in private sector (Table 2).The distribution of cases exemplifies the complexity of the health system and potential associated challenges for patients since the private and public sectors are not interconnected.Moreover, within the public sector there is no processes to reference patients, making challenging to have opportune diagnosis, treatment, and follow-up.
According to the survey applied to AEP, emergency care in medical units, was granted to 33% of patients, being 67.3% men and 32.7% women.The average number of visits was 7.6/patient-year (range 1-12/patient-year).They represented 61.2% of all acute attack care.The rest of the acute attacks of minor severity, were attended at home (13.5%), and in outpatient consultation (25.3%) (Table 2).Regarding findings suggesting HAE, AEP reported the presence of repeated angioedema, with approximately 40-50% of cases involving abdominal pain, lack of response to antihistamines and corticosteroids in 90-95%, family history (76% public sector, 90% private sector), and beginning of symptoms in childhood or adolescence (70% public sector, 80% private sector).From a clinical perspective, the main prodrome were: abdominal pain with associated gastrointestinal symptoms (diarrhea and nauseavomiting) 47.9% and serpiginous erythema 38.0%, continuing during the clinical manifestation.The main manifestations were in terms of frequency: skin edema 67.5%, abdominal pain without gastrointestinal symptoms 42.1%, laryngeal edema 27.5%, and other symptoms 2% (Fig. 2).The key diagnostic studies carried out in these patients corresponded to the determination of C1-INH levels, C1-INH functional quantification, C1q, C3, and C4 levels (36% in public sector and 50% in private sector).In cases clinically suspected to be acquired, anti-C1INH autoantibodies were determined (15% in both sectors).
Genetic study was carried out in a small group of cases (10%), being pathogenic variants in the SERPING1 gene the finding in 74% of the patients with genetic test.Regarding acute attack frequency of 12.3 per capita per year (range 4.0-27.1) is mentioned by the AEP (Table 1), being more frequent after adolescence with around 60% of the events in those over 17 years of age (Data not shown).
According to the survey responses analysis to the therapeutic approach, fresh frozen plasma was the most used medical treatment for the acute attack in both public and private sectors, followed by icatibant and C1-INH concentrate.On the other hand, danazol was the most used treatment for STP (Table 3).

DISCUSSION
Mexican nationwide HAE prevalence for 2019 was estimated to be of 2024 total cases, corresponding to a rate of 0.805/50 000 inhabitants (CI 95% 0.805-0.808),with an estimated of 376 patients identified as C1-INH deficiency.
The results obtained with this methodology show consistency with previously published data.Banerji et al 13 (2020) reported 78.4% HAE type 1 in the United States US) population.In the Austrian registry 33 80.2% and in Brazilian cohort 34 76.1% were HAE type 1. Comparatively, in our study corresponds to 90.1% (n ¼ 317) HAE Type 1 in estimated total population prevalence.Regarding the presence of a family history, Banerji et al 13 found it in 78.4% of their population.In this study it was estimated in 75.0%, in the Austrian registry 33 83.7%, and in Brazilian cohort 86.9%. 34Regarding the onset of symptoms in these 4 populations it occurred mainly in childhood-adolescence. 13,33,34nsidering the severity of the attacks, in our results from the Mexican and Brazilian populations, around 75% of the events were moderate to severe, with the difference that in Mexico 50% of the cases corresponded to moderate, while in Brazil just over 50% were serious. 34In general terms, a similar pattern is observed in various countries, with the sum of moderate plus severe attacks dominating.That implicitly shows that the problem is not close to being controlled.The enormous burden of disease showed by Banerji et al 13 for the US population allows us to see this clearly: Anxiety and depression were frequently identified, affecting 35.3% and 20.9% of patients, respectively. 13  describing this type of implication should be addressed.
Regarding the number of attacks, Banerji et al 13 found for a period of 6 months prior to the application of a specific survey, an average of 11.1 events per capita in the US population, similar to our results (12.3 average events/patient-year).Also, Banerji et al found that most patients (78.7%) reported an attack in the last month and that 41.8% suffered an attack within the previous week of the survey's application. 13e high number of attacks in frequency, together with their duration for certain patients (33.9% suffered attacks lasting 2 days), should be further evaluated since it correlates with negative effects on QoL, as well as on the mental sphere. 12t certainly represents a clear area of opportunity to incorporate human development dimensions (QoL, days of productivity lost, mental distress, among others) to HAE research.
The NHS share for public institutions (70.7%,Table 2) represents a big challenge in Mexico because the gross domestic product assigned to health attention is below the recommended by the OCDE.(5.4 vs 8.8, respectively 2021). 35e main strength of the study was identified in the methodological approach with the use of mixed methods (carried out in 4 phases: a systematic review of the literature and meta-analysis; review of the NHS databases and systematic reports; Delphi panel and survey; and an epidemiologic model).This allowed the prevalence of HAE in Mexico to be estimated for the total population. 27,28Likewise, by obtaining and adding the records coded by D84.1 of the ICD 10 of all the different databases of the most relevant public institutions and the private sector of the NHS (1/1/2019-31/12/2019), [19][20][21][22][23] it was possible to identify the number of cases seen and treated in the NHS during 2019, as well as the corresponding distribution by gender and age group, both for outpatients and inpatients.The clinical profile and therapeutic behavior obtained from the ad hoc survey results applied to 40 Mexican AEP gave specific details on HAE diagnosis and treatment as real-world experience.
It is worthy to mention that our study has intrinsic limitations regarding the use of administrative databases, such as the lack of clinical, socioeconomic, and QoL variables, a fact that imposes the need for using other sources to correct the information gap, thus implicating the additional risk of introducing bias in the interpretation of information coming from different sources.Likewise, it is important to keep in mind that the registry encoded by ICD 10 D84.1 identifies only cases with C1-INH deficiency but does not differentiate between Type 1 or 2, and of inherited and acquired cases.In relation to the above, in the present investigation, we estimated HAE cases through data from the survey administered to AEP.According to this, the total number of records considered HAE was 352 (96.3%), 317 were estimated as HAE Type 1 (90.1% [95% CI 84.3%-91.1%]),and 35 cases were considered HAE Type 2 (9.9% [95% CI 9.2%-10.4%]).In recent international literature, similar criteria are used to filter or differentiate between hereditary and acquired cases. 36,37mmarizing, it is known that EMR (electronic medical record) databases for investigating HAE in retrospective models can be challenging due to the lack of specific diagnostic codes for this condition, combined with the frequency of delayed diagnosis and/or misdiagnosis. 38In this regard, there are structured EMRs such as ICD 10 D84.1, and unstructured ones such as medical notes. 38iven that HAE is a rare disease, being able to count on both categories of information, elements that provide more specific data for a more accurate diagnosis that make it possible to estimate hereditary cases from early stages due to C1-INH deficiency qualitatively or quantitatively, or with normal C1 esterase inhibitor, will undoubtedly be key elements for a better understanding and management of this clinical condition.In addition to testimonial information from AEP could and/or considered to add value. 38spite the recognized usefulness of structured EMRs, there are recent reports that show underrepresentation in the number of patients. 38ikewise, the value of adding more sources of information has also been documented. 38In this study, the addition of the survey supports the results.Above all, relying just in ICD 10 D84.1 could be a risky situation, since this code is just related to C1-INH deficiency, thus, leaving out even the chance of a proxy for estimate HAE nC1-INH, the estimation with the present methodology accounts just for 90%-95% of all HAE cases.
On the other hand, administrative databases have many advantages in an environment where health data are scarce (such as Mexico) as a valuable, standardized, and systematic resource.Although our methodology to estimate prevalence and frequency proxies is not a substitute for other designs (population-based epidemiological studies, an accurate and systematic National HAE registry), it is less expensive (economically and in run time) and more convenient than prospective follow-up studies due to the difficult of enrolling significant amounts of patients in a reasonable period.Future designs in settings where electronic comprehensive health records are available, could include detailed information on treatment and clinical manifestations per capita.
These results are considered to provide an interesting and innovative approach in the Mexican level to estimate and understand the calculated burden of HAE in Mexico, as well as an additional step in understanding the diagnose and treatment profile of the disease locally.
The current ICD 10 version for 2019 considered the ICD 10 code D84.1 to describe cases with C1-INH deficiency. 39Although there are other complement system deficiencies documented and considered by this ICD 10 code, as shown in the specialized literature 40 other different to C1 inhibitor deficiency, account for less than 15% (and individually being very rare), when using the mapping for the determination of the applicable DRG, the meaning most frequently found when using ICD 10 D84.1, substantially, corresponds to its meaning as C1-INH deficiency.Therefore, and according to the panel of AEP, as well as some regional clinical practice guidelines and the health information system of the Mexican MoH, this code is considered an adequate proxy to identify and follow up on cases of C1-INH deficiency, whether acquired or hereditary and of these Type 1 or Type 2. This coding does not allow to directly differentiate between HAE Type 1 and Type 2 cases, as well as between hereditary and acquired cases of angioedema, for this, other sources of information and health modeling must be used.
In support of consideration and use as a proxy for C1-INH deficiency analysis, it is known that this ICD 10 code, in practice (diagnostic coding) is related to the following ORPHANET codes: ORPHA 528663 (AE acquired); ORPHA 528623 (hereditary AE), ORPHA 100050 (HAE Type 1), ORPHA 100051 (HAE Type 2), and with MS-DRG 40.0.
With the advent of ICD 11, it is now possible to differentiate directly from the coding, between cases of AE specifically (no longer including the code to other complement deficiencies).The specific ICD 11 codes are ICD 11 4A 00.14 (HAE) and ICD 11 4A 00.15 (AAE).Despite this, in order to discriminate between the different types of HAE, it will still be necessary to use complementary methods such as medical notes and surveys with treating physicians.This implies knowing firsthand, at least the number of cases of HAE and AAE.

CONCLUSIONS
The present study represents the first publication that provides an estimate of the number of patients with HAE (focusing on Type 1 which represents more than 90% of the cases), by gender and age in Mexico based largely on local data sources.It also includes data on epidemiology, diagnosis, presentation, attack characteristics, healthcare processes, and treatment patterns to increase the understanding of HAE in the Mexican population.It is important to note that according to this study, the recognized and treated cases of HAE in the NHS in a full year (2019) represent only 16.6% of the estimated prevalence for 2019.

Table 1 .
Hereditary Mexican Social Security Institute), 41.9% (114) in the SSA Angioedema: Estimated prevalence.Cases by gender.Cases by Type.Acute Attack by average severity.Acute Attack by annual average per capita.Specific diagnosis rate (ICD 10: D84.1) in Mexico, 2019

Table 2 .
(Continued) Hereditary Angioedema (ICD 10 D84.1 defects in the complement system): Outpatient care, inpatient care, hospital cases and main causes of hospitalization.Distribution according to age groups and Institutions in Mexico, 2019 IMSS, Instituto Mexicano del Seguro Social; SSA, Secretaría de Salud; ISSSTE, Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado.
Considering these results and the psychological impact on patients, future studies