Survivors

A frailty index counts deficits in health. These deficits were defined as symptoms, signs, disabilities and diseases [5 ]. All health deficits, including continuous, ordinal and binary variables, were taken from the PEP survey data dictionary. Restricted activity, disability in Activities Daily Living (ADL) and Instrumental ADL, impairments in general cognition and physical performance (e.g. impaired grip strength, impaired walking), co-morbidity, self-rated health, and depression/mood were evaluated.
Variables can be included in a frailty index if they satisfy the following 5 criteria:
1) The variables must be deficits associated with health status. Attributes such as graying hair, while age-related, are attributes and therefore not included. 2) A deficit's prevalence must generally increase with age, although some clearly age-related adverse conditions can decrease in prevalence at very advanced ages due to survivor effects. 3) Similarly, the chosen deficits must not saturate too early. For instance, age-related lens changes resulting in problems with accommodation (presbyopia) are nearly universal by age 55; in other words, as a variable, presbyopia saturates too early to be considered as a deficit here. 4) When considering the candidate deficits as a group, the deficits that make up a frailty index must cover a range of systems – if all variables were related to cognition, for example, the resulting index might well describe changes in cognition over time, but would be a cognitive impairment index [18 (link)] not a frailty index. 5) If a single frailty index is to be used serially on the same people, the items that make up the frailty index need to be the same from one iteration to the next [19 (link)]. The requirement to use the same items need not apply to comparisons between samples – i.e. samples that use difference frailty indexes appear to yield similar results [5 ].
Deficits should be added until there are at least 30–40 total deficits. There needs to be a minimum number of deficits. In general, the more variables that are included in a frailty index, the more precise estimates become. Similarly, estimates are unstable when the number of deficits is small – about 10 or less. Even so, an index with 30–40 variables has been shown to be sufficiently accurate for predicting adverse outcomes [6 (link),14 (link)]. Furthermore, a frailty index can be constructed using information that is readily available in most health surveys, and is clinically tractable – i.e. it uses an amount that would be gathered in many routine health assessments of older adults [5 ].

We developed the SURVIVOR tool kit for assessing SVs for short-read data that contains several modules. The first module simulates SVs given a reference genome file (fasta) and the number and size ranges for each SV (insertions, deletions, duplications, inversions and translocations). After reading in the reference genome, SURVIVOR randomly selects the locations and size of SV following the provided parameters. Subsequently, SURVIVOR alters the reference genome accordingly and prints the so altered genome. In addition, SURVIVOR provides an extended bed file to report the locations of the simulated SVs.
The second module evaluates SV calls based on a variant call format (VCF) file56 (link) and any known list of SVs. A SV was identified as correct if (i) they were of same type (for example, deletion); (ii) they were reported on same chromosome and (iii) the start and stop coordinates of the simulated and identified SV were within 1 kb (user definable).
The third module of SURVIVOR was used to filter and combine the calls from three VCF files. In our case, these files were the results of DELLY, LUMPY and Pindel. This module includes methods to convert the method-specific output formats to a VCF format. SVs were filtered out if they were unique to one of the three VCF files. Two SVs were defined as overlapping if they occur on the same chromosome, their start and stop coordinates were within 1 kb, and they were of the same type. In the end, SURVIVOR produced one VCF file containing the so filtered calls. SURVIVOR is available at github.com/fritzsedlazeck/SURVIVOR.

The annual incidence rate for CHD (8 (link)) and stroke was calculated from the Framingham study equations (9 (link)). As four-index cohorts had been defined with specific characteristics, there was a need to calculate the risk based on those profiles. Based on the literature, one out of four CHD events are fatal in the first year (10 (link)), while 60% of them are pre-hospital deaths (11 (link)). Therefore, it was assumed that of those who have a CHD event in the model, 25% die in the first year. Approximately 60% of these deaths were costless as they are pre-hospital deaths. Regarding first-year stroke mortality, the range varies in different resources and is reported from 22 to 34% (12 (link)). For this analysis, the rate was applied from the largest available cohort (13 (link)). Almost 25% of stroke events are fatal in the first year, while half of them occur during the first 28 days. Therefore, it is assumed that although at the end of the cycle, they move to the death state, 40% of the cycle cost should be considered for them. The fatality rate for stroke and CHD survivors was derived from a study that had been done on the Iranian population (14 (link)). The background mortality rate from all causes other than stroke and CHD is calculated by excluding the total death attributed to these two diseases from the Iranian life table¹. The total mortality rate of these two events had been calculated in Tehran Lipid and Glucose Study (TLGS). At first, the annual rates were derived from the life table and then the CHD- and stroke-attributed deaths were excluded.

Two separate Markov decision models were developed to compare the long-term costs and health benefits of the IraPEN program (primary CVD prevention) with the status quo (no prevention) in two distinct scenarios. In the base case scenario, individuals without diabetes were included, while patients with diabetes were included in the alternative scenario. Each Markov model has four health states with transitions between the states according to age, sex, and the CVD risk characteristics of participants (Figure 1). In contrast to the usual Markov models, which are structured based on cohorts with average profiles, we decided to categorize the individuals based on their CVD risks. As the intervention (treatment) varied according to CVD risk level, it is logical to model them separately. In this way, we can take into account their specific characteristics. Therefore, based on WHO/ISH CVD risk prediction charts for EMR B, four index cohorts were constructed (5 ). These hypothetical cohorts were used as a representative for individuals with low, moderate, high, and very high CVD risk profiles. The CVD risk state represents the starting point for all people who are 40 years old. It was assumed that people in this state may either remain in the same health state, move to the stroke state, or CHD (coronary heart disease) state, or die. As long as they are event-free, these individuals can stay in a healthy state, but after the first event, they move to the CHD or stroke state and stay there until their death.
In WHO/ISH CVD risk prediction charts, the CVD risk is calculated based on individuals' age and risk factors such as blood pressure, lipid profile, diabetes, and smoking status and categorized into the following five groups: below 10% (low-risk group), between 10 and 19% (moderate-risk group), between 20 and 29% (high-risk group), between 30 and 39%, and above 40% (very high-risk group). As the individuals in the two latter groups are treated the same, in the IraPEN program, whoever has a CVD risk above 30% is categorized as the very high-risk group.
Therefore, considering what was mentioned earlier, all the Iranians aged older than 40 years who did not have CHD or stroke events before were eligible for this program. According to the recent census (2016), 31.16% of Iranians were older than 40 years (6 ). By adding individuals aged older than 30 years with the aforementioned risk factors, we can conclude that this program is going to screen at least 25 million people yearly.
The healthcare perspective and a 40-year time horizon were adopted for this analysis. As the analysis is a comparison between IraPEN (intervention) and status quo (no intervention) which both have the same Markov structure and transition probabilities, it is not expected that half cycle correction (HCC) approach makes any difference in ICER results; therefore, HCC was not applied to this analysis (7 (link)).
The hypothetical cohorts were used as a representative for individuals with low, moderate, high, and very high CVD risk profiles (Table 1). Progressively, a proportion of the cohort can go to the CHD state, who are the survivors of the first CHD event, or to the stroke state who are the survivors of the first stroke event. Those CHD and stroke events that were fatal moved to the death state. In general, the people in these two states are at a higher risk of dying from CHD or stroke, but they may die from any other causes like the normal population. Table 2 summarizes the assumptions of this analysis.

A healthcare perspective was adopted; therefore, we only included costs associated with healthcare such as direct medical costs (Table 4). The costs considered in the model are the cost of IraPEN screening, the cost of IraPEN monitoring, the cost of CHD survivors, and the cost of stroke survivors. It is assumed that the cost of individuals who are event-free in the status quo is zero as long as undiagnosed or untreated. These two facts were considered for the status quo costs. Furthermore, it was assumed that the cost of dying was equal to zero. According to PEN protocols, the needed resources for each index cohort were identified. Then, the items were quantified based on discussions with the physicians and supervisors of the visited centers. The cost of index cohorts consists of two different types. First, variable costs are different for each group based on the characteristics of each. Second, fixed cost is the same for all and consists of staff training, administration, IT, promotional stuff, and leaflets. The unit price of each item was derived from the last report of the Ministry of health (23 ). The report estimated all the costs related to IraPEN implementation except the medications. In addition, the reported costs were adjusted by the 2018 inflation rate and the cost of each cohort was calculated.
The cost of CHD state and stroke state was derived from an Iranian CE that had estimated the cost of these two states (24 (link)). These two costs contain all the related medical costs such as hospital admissions and procedures, monitoring, follow-ups, medications, and secondary prevention (Table 5). Based on experts' opinions, it is assumed that the cost of CHD after the first year would be a third and the cost of stroke state after the first year would be a quarter. In addition, it is assumed that the standard error of costs for the consecutive year is 10% of the mean.