lib/health-data-standards/models/cqm/measure.rb
module HealthDataStandards
module CQM
class Measure
include Mongoid::Document
include Mongoid::Timestamps
MSRPOPL = 'MSRPOPL'
store_in collection: 'measures'
field :id, as: :id, type: String
field :sub_id, type: String
field :cms_id, type: String
field :name, type: String
field :title, type: String
field :description, type: String
field :subtitle, type: String
field :short_subtitle, type: String
field :hqmf_id, type: String
field :hqmf_set_id, type: String
field :hqmf_version_number, type: String
field :nqf_id, type: String
field :type, type: String
field :category, type: String
field :population_ids , type: Hash
field :oids, type: Array
field :population_criteria, type: Hash
field :data_criteria, type: Array, default: []
field :source_data_criteria, type: Hash, default: {}
field :measure_period, type: Hash
field :measure_attributes, type: Hash
field :populations, type: Array
field :preconditions, type: Hash
field :hqmf_document, type: Hash
field :map_fn, type: String
field :continuous_variable, type: Boolean
field :episode_of_care, type: Boolean
field :aggregator, type: String
embeds_many :prefilters
scope :top_level_by_type , ->(type){where({"type"=> type}).any_of({"sub_id" => nil}, {"sub_id" => "a"})}
scope :top_level , ->{any_of({"sub_id" => nil}, {"sub_id" => "a"})}
scope :order_by_id_sub_id, ->{order_by([["id", :asc],["sub_id", :asc]])}
index oids: 1
index hqmf_id: 1
index category: 1
index sub_id: 1
index _id: 1, sub_id: 1
index({bundle_id: 1, hqmf_id: 1, sub_id: 1})
validates_presence_of :id
validates_presence_of :name
def self.categories(measure_properties = [])
measure_properties = Array(measure_properties).map(&:to_s) | %w(
name description nqf_id cms_id hqmf_id continuous_variable episode_of_care
)
pipeline = []
pipeline << {'$group' => measure_properties.inject({
'_id' => "$id",
'subs' => {'$push' => {"sub_id" => "$sub_id", "short_subtitle" => "$short_subtitle"}},
'sub_ids' => {'$push' => "$sub_id"},
'category' => {'$first' => "$category"}
}) do |h, prop|
h[prop] = {"$first" => "$#{prop}"}
h
end
}
pipeline << {'$group' => {
_id: "$category",
measures: {
'$push' => measure_properties.inject({
'id' => "$_id",
'hqmf_id' => "$_id",
'subs' => "$subs",
'sub_ids' => "$sub_ids"
}) do |h, prop|
h[prop] = "$#{prop}"
h
end
}
}}
pipeline << {'$project' => {'category' => '$_id', 'measures' => 1, '_id' => 0}}
pipeline << {'$sort' => {"category" => 1}}
Mongoid.default_session.command(aggregate: 'measures', pipeline: pipeline)['result']
end
# Returns the hqmf-parser's ruby implementation of an HQMF document.
# Rebuild from population_criteria, data_criteria, and measure_period JSON
def as_hqmf_model
@hqmf ||= HQMF::Document.from_json(self.hqmf_document)
end
def key
"#{self['id']}#{sub_id}"
end
def is_cv?
! population_ids[MSRPOPL].nil?
end
def self.installed
Measure.order_by([["id", :asc],["sub_id", :asc]]).to_a
end
# Finds all measures and groups the sub measures
# @return Array - This returns an Array of Hashes. Each Hash will represent a top level measure with an ID, name, and category.
# It will also have an array called subs containing hashes with an ID and name for each sub-measure.
def self.all_by_measure
reduce = 'function(obj,prev) {
if (obj.sub_id != null)
prev.subs.push({id : obj.id + obj.sub_id, name : obj.subtitle});
}'
self.moped_session.command( :group=> {:ns=>"measures", :key => {:id=>1, :name=>1, :category=>1}, :initial => {:subs => []}, "$reduce" => reduce})["retval"]
end
def display_name
"#{self['cms_id']}/#{self['nqf_id']} - #{name}"
end
def set_id
self.hqmf_set_id
end
def measure_id
self['id']
end
def continuous?
population_ids[MSRPOPL]
end
def title
self.name
end
def all_data_criteria
return @crit if @crit
@crit = []
self.data_criteria.each do |dc|
dc.each_pair do |k,v|
@crit << HQMF::DataCriteria.from_json(k,v)
end
end
@crit
end
# Builds the query hash to pass to MongoDB
# Calling this method will create Prefilters if they do not exist on the
# measure
def prefilter_query!(effective_time)
self.build_pre_filters! if self.prefilters.empty?
if self.prefilters.count == 1
self.prefilters.first.build_query_hash(effective_time)
else
self.prefilters.inject({}) do |query, pf|
query.merge(pf.build_query_hash(effective_time)) do |key, new_val, old_val|
new_val.merge(old_val)
end
end
end
end
# For submeasures, this will return something like IPP_1
def ipp_id
ipp_hqmf_id = self.population_ids['IPP']
pop_id, pop_criteria = hqmf_document['population_criteria'].find do |population_id, population_criteria|
population_criteria['hqmf_id'] == ipp_hqmf_id
end
pop_id
end
def build_pre_filters!
dc = self.data_criteria.inject({}) do |all_dc, single_dc|
key = single_dc.keys.first
value = single_dc.values.first
all_dc[key] = value
all_dc
end
dc.each_pair do |criteria_name, data_criteria|
if data_criteria['definition'] == 'patient_characteristic_birthdate'
if data_criteria_in_population?(self.ipp_id, criteria_name)
prefilter = Prefilter.new(record_field: 'birthdate',
effective_time_based: true)
if data_criteria['temporal_references']
data_criteria['temporal_references'].each do |tr|
if tr['type'] == 'SBS' && tr['reference'] == 'MeasurePeriod'
years = nil
if tr['range']['high']
prefilter.comparison = '$gte'
years = tr['range']['high']['value'].to_i
elsif tr['range']['low']
prefilter.comparison = '$lte'
years = tr['range']['low']['value'].to_i
end
prefilter.effective_time_offset = 1 + years
self.prefilters << prefilter
end
end
end
end
end
end
end
private
def data_criteria_in_population?(population_id, criteria_name)
criteria_in_precondition?(self.hqmf_document['population_criteria'][population_id]['preconditions'], criteria_name)
end
def criteria_in_precondition?(preconditions, criteria_name)
preconditions.any? do |precondition|
(precondition['reference'] == criteria_name) ||
(precondition['preconditions'] && criteria_in_precondition?(precondition['preconditions'], criteria_name))
end
end
end
end
end