JP2005000265A

JP2005000265A - Method for building health conditions-specific onset risk knowledge and health management equipment

Info

Publication number: JP2005000265A
Application number: JP2003164519A
Authority: JP
Inventors: Mariko Yamamoto; 真理子山本; Hideyuki Ban; 伴　　秀行; Takanobu Osaki; 高伸大崎; Kazuyuki Shimada; 和之島田; Kei Masuda; 圭増田
Original assignee: Hitachi Ltd
Current assignee: Hitachi Ltd
Priority date: 2003-06-10
Filing date: 2003-06-10
Publication date: 2005-01-06

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for building a health condition-specific onset risk knowledge by accurately providing differences in onset risk. <P>SOLUTION: The method has the process (S120) of defining health conditions aimed at, setting conditions for extracting accumulated health and contraction data in the health conditions aimed at and extracting accumulated health and contraction data meeting the extraction conditions, the process (S130) of computing an age-specific incidence Ri (ti) for an age ti, and the process (S140) of judging if the number at a point Ri (ti) used for regression analysis is sufficient or not. Furthermore, the method has the process of repeating S120 and S130 until the number becomes sufficient, the process (S150) of computing the incidence R (t) for all ages by obtaining parameters A, B and M from Ri (ti) by regression analysis showing the incidence after passing the judgment S140, the process (S160) of computing a health lifetime K by using R(t) and the process (S170) of writing the extraction conditions and the health lifetime K in the health condition-specific onset risk knowledge. The accurate computation of an incidence or an occurrence from limited data is possible. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

【０００１】
【発明の属する技術分野】
健康状態と生活習慣病の発症リスクとの対応知識を構築する健康状態別の発症リスク知識構築方法及び健康状態別の発症リスク知識を活用して健康管理に寄与する健康管理装置に関する。
【０００２】
【従来の技術】
健康管理装置においては、健康管理の動機付けを行うため、個人の特性を考慮することが重要である。従来、健康状態と健康度や発症リスクの対応知識を用いて健康管理に寄与する健康状態別の発症リスク知識構築方法及び健康管理装置に関する報告がある（例えば、特許文献１、特許文献２）。特許文献１では、多数のデータから疾病モデルと標準モデルを作成し、２つのモデルと個人の健診データを比較して個人の健康度を評価している。特許文献２では、薬物介入大規模試験の結果や、疫学調査結果により導出された方程式など、一定のルールを用いて健康データを発症リスクへ変換している。
【０００３】
【特許文献１】
特開２００２−６３２７８号公報
【特許文献２】
特開２００２−２４４０１号公報
【０００４】
【発明が解決しようとする課題】
上記の従来技術では、個人の特性を考慮した健康度又は発症リスクを提供できる。しかし、少数のデータから正確に発症リスクを算出できないため、健康データに健康度や発症リスクを対応させる知識が、健康度又は発症リスクが提供される利用者の地域性など、全国平均と異なる原因となる因子を反映することが難しいという問題と、健康状態のわずかな違いに対して正確に発症リスクの違いを求めることが難しいという問題の２つの問題があった。本発明はこのような事情を考慮してなされたものである。
【０００５】
本発明の目的は、利用者の地域性などの全国平均と異なる原因となる因子を反映し、さらに、健康状態のわずかな違いに対しても正確に発症リスクの違いを求められる、健康状態別の発症リスク知識構築方法及び健康管理装置を提供することにある。
【０００６】
【課題を解決するための手段】
（１）本発明の健康状態別の発症リスク知識構築方法は、健康状態と生活習慣病の発症リスクとの対応知識を構築する健康状態別の発症リスク知識構築方法において、以下の特徴を有している。性別や生年月日などの発症リスクに関わり生涯不変な基本情報である個人基本情報と、血圧や体重などの現在の健康状態を機器等を用いて定量的に測定した結果である検査結果と、食事内容やストレスなどの現在の生活習慣や健康状態を質問用紙記入や問診等を用いて収集した結果である問診結果と、現在の疾病の有無及び疾病名である罹患情報とを用いて、個人基本情報及び検査結果及び問診結果と罹患情報との関連性を個人別に分析して、具体的には、発症に関する知見を定式化した回帰式を用いて、罹患率又は発症率を算出する罹患率・発症率回帰ステップを有すること。
【０００７】
（２）本発明の健康管理装置は、以下の特徴を有する。健康状態と生活習慣病の発症リスクとの対応知識を活用して健康管理に寄与する健康管理装置において、性別や生年月日などの発症リスクに関わり生涯不変な基本情報である個人基本情報と、血圧や体重などの現在の健康状態を機器等を用いて測定した結果である検査結果と、食事内容やストレスなどの現在の生活主観や健康状態を質問用紙記入や問診等を用いて収集した結果である問診結果と、現在の疾病の有無及び疾病名である罹患情報とを用いて、個人基本情報と検査結果と問診結果と罹患又は発症率との関係を算出する罹患率・発症率回帰部を有している。罹患率・発症率回帰部は、個人基本情報及び検査結果及び問診結果と罹患情報との関連性を個人別に分析して、具体的には、発症に関する知見を定式化した回帰式を用いて、罹患率又は発症率を算出する。
【０００８】
上記（１）及び（２）において、以下の特徴を有する。
（イ）回帰式は、中年、老年など年齢区分に対応する複数の項を有し、各項のべき数がパラメータであるか、あるいはデータを元にして決定した定数である。
（ロ）Ｌ（ｔ）を年齢ｔでの生存率、α１、α２を、例えば、１００など、生存可能な年齢の上限値、Ａ、Ｂ、Ｍをパラメータとするとき、回帰式は（数１）により表れされ、年齢ｔでの罹患率Ｒ（ｔ）を表す。
【０００９】
【数１】

【００１０】
発症に関する知見を定式化した回帰式の使用により、少数のデータからも正確に罹患率又は発症率が算出可能になり、地域別あるいは健康状態別に分類して少数になったデータに対しても正確な罹患率あるいは発症率を算出可能になり、健康状態や地域性と発症リスクの詳細で正確な関係を得ることができる。この結果、健康状態別の発症リスク知識構築方法及び健康状態別の発症リスク知識を活用して健康管理に寄与する健康管理支援装置を提供できる。
【００１１】
【発明の実施の形態】
本発明の健康状態別の発症リスク知識構築手方法は、性別や生年月日などの発症リスクに関わり生涯不変な基本情報である個人基本情報と、血圧や体重などの現在の健康状態を機器等を用いて定量的に測定した結果である検査結果と、食事内容やストレスなどの現在の生活習慣や健康状態を質問用紙記入や問診等を用いて収集した結果である問診結果と、現在の疾病の有無及び疾病名である罹患情報とを含む健康・罹患データを多数蓄積した蓄積健康・罹患データから、個人基本情報及び検査結果及び問診結果と罹患情報との関連性を個人別に分析して、罹患率など、発症リスクを算出する健康状態別の発症リスク知識構築方法であり、発症に関する知見を定式化した回帰式を用いて、罹患率又は発症率を算出する罹患率・発症率回帰ステップを有する。
【００１２】
本発明の健康管理装置は、健康状態と生活習慣病の発症リスクとの対応知識を活用して健康管理に寄与する健康管理装置であって、上記の発症リスク知識構築方法を用いる。発症リスク知識の構築に用いるデータは装置を運用する施設に蓄積した健康・罹患データ又は装置の運用者が設定したデータであり、これらのデータを用いて上記の発症リスク知識構築方法により構築した健康状態別の発症リスク知識を施設独自の健康状態別の発症リスク知識として用いる。本発明の装置は、個人の健康・罹患データを入力すると、健康状態別の発症リスク知識を検索し、入力した健康・罹患データに対する発症リスクを取得して出力する。なお、本発明の健康管理装置の入力データは現在の罹患情報を含まない他は健康・罹患データと同様のデータである健康データでもよい。
【００１３】
以下、本発明の発症リスク知識構築方法の一例を図面を用いて詳細に説明する。
（実施例１）：健康状態別の発症リスク知識構築方法に関する説明。
【００１４】
以下の説明では、説明を容易にするため、発症リスクは糖尿病の健康寿命Ｈ（ｔ）で表す。糖尿病の健康寿命とは糖尿病を発症するまでの平均年数であり、糖尿病の罹患率から（数２）によって算出される量である。
【００１５】
【数２】

【００１６】
（数２）で、ｔは年齢、Ｋ（ｔ）は年齢ｔにおける健康寿命、Ｌ（ｔ’）は年齢ｔ’における生存率、Ｒ（ｔ’）は年齢ｔ’における罹患率（（数１）により示される）を表し、αは、例えば、１００など、生存可能な年齢の上限値である。
【００１７】
図２は、本発明のの実施例の発症リスク知識構築方法で使用される蓄積データの一例を示す図である。
【００１８】
図２（ａ）は、個人の基本情報や健康状態と罹患情報を表す、蓄積健康・罹患データであり、性別、年齢など個人の基本情報（個人の属性を表す情報）２１０、検査日や検査施設などの検査基本情報（検査属性）２２０、身長、血圧、生理検査値などの検査結果（検査値の情報）２３０、生活習慣や既往歴、家族歴などの問診結果（問診の情報）２４０と、現在の罹患状態の情報２５０を含む。蓄積健康・罹患データは、通し番号である検査特定ＩＤ（２１２）として、例えば、“Ｎ００００００１”を、個人特定ＩＤ（２１１）として、例えば、“Ｐ００００００１”を付けて管理されており、検査特定ＩＤにより一意に決定できる。個人特定ＩＤと検査特定ＩＤの関係は１対多対応である。即ち、個人特定ＩＤ“Ｐ００００００１”のデータのうち検査特定ＩＤが“Ｎ００００００１”と“Ｎ００００００２”の２つのデータがある。この１対多対応は、１人の人が複数回検査を受けた結果をも蓄積している結果である。
【００１９】
図２（ｂ）は、蓄積健康・罹患データを健康状態別に分類するため、又は、着目する健康状態にある健康・罹患データを抽出するための抽出条件であり、蓄積健康・罹患データの項目を用いて、例えば、個人の基本情報、検査結果、問診結果、各項目についての命題２６０、２７０、２８０の論理積で作成される。命題２６０は個人の属性を表す項目についての命題、命題２７０は検査値の情報の項目についての命題、命題２８０は問診の情報の項目についての命題を示す。抽出条件として、例えば、性別が男性、かつ、年齢が２０歳以上３０歳以下、かつ、脈拍が５０以下、かつ、血糖値が９０以下かつ１週間の運動量が２０００ｋｃａｌ以上を一例とする条件を作成し、通し番号である抽出条件特定ＩＤ（２２２）として、例えば、“Ｃ００００００１”を付けて管理している。
【００２０】
図２（ｃ）は健康状態と発症リスクの関係、即ち、抽出条件と健康寿命の対応関係を保持する健康状態別の発症リスク知識であり、健康状態を表す情報２９０、例えば、抽出条件の管理番号と、健康寿命２９５を含む。例えば、条件Ｃ００００００１に従って検査特定ＩＤが“Ｎ００００００３”のデータや検査特定ＩＤが“Ｎ００００００４”のデータなど、多数のデータを抽出し、抽出したデータから、（数３）によって、罹患情報／糖尿病の欄が“治療中”のデータと“健康”のデータをカウントして年齢ｔｉ＝２５における年齢別罹患率Ｒｉ（２５）を求め、条件Ｃ００００００２、Ｃ００００００３に従って同様に求めたＲｉ（３５）、Ｒｉ（４５）とを合わせた３点を、（数１）により回帰して、例えば、パラメータＡ＝０．３、Ｂ＝０．１、Ｍ＝２．２を得た後、（数２）により健康寿命を算出した結果を“５５”、“６８”、“７９”と得ている。さらに同様に、他の健康状態を表す抽出条件“Ｃ００００００４”以降を設定して、健康寿命を“５１”…と得ている。
【００２１】
【数３】

【００２２】
ここで、（数１）、（数２）の年齢ｔと罹患率Ｒが添え字を持たず、（数３）の年齢ｔｉと罹患率Ｒｉが添え字ｉを持つのは、前者が全ての年齢で値を持つのに対し、後者は回帰に用いるデータの数Ｎだけ値を持つためである。ｔｉは（数３）の右辺に用いたデータの最大年齢と最小年齢の中間値、即ち、年齢階級の代表値である。なお、（数１）、（数２）で用いる年齢別生存率Ｌ（ｔ）は、何らかの手段により算出された、健康寿命を提供する集団の死亡率でもよいし、平均的な値、例えば、厚生労働省の生命表から算出した値でもよい。
【００２３】
図１は、本発明の実施例の発症リスク知識構築方法の処理フローの一例を示す図である。
【００２４】
処理が開始されると（Ｓ１１０）、“２０歳代の男性で血糖値が９０未満”など、着目する健康状態を定義して、着目した健康状態にある蓄積健康・罹患データを抽出する抽出条件を設定し、この抽出条件を満たす蓄積健康・罹患データを抽出する（Ｓ１２０）。次に、年齢ｔｉに対する年齢別罹患率Ｒｉ（ｔｉ）を（数３）により算出する（Ｓ１３０）。次に、回帰分析に用いる点Ｒｉ（ｔｉ）の数が充分であるかを判定し（Ｓ１４０）、充分になるまでＳ１２０とＳ１３０を繰り返す。このループは、回帰を行うには一般的に数個の点が必要であるため設けたもので、例えば、“２０歳代の男性で血糖値が９０未満”、“３０歳代の男性で血糖値が９０未満”、“４０歳代の男性で血糖値が９０未満”にあたる年齢別罹患率Ｒ（２５）、Ｒ（３５）、Ｒ（４５）を算出する。
【００２５】
判定Ｓ１４０を通過すると、（数１）により罹患率を表す回帰分析によりＲｉ（ｔｉ）からパラメータＡ、Ｂ、Ｍを求めて、全年齢での罹患率Ｒ（ｔ）を算出可能にし、健康寿命Ｋを算出するのに必要な罹患率Ｒ（ｔ）を算出する（Ｓ１５０、なお、このステップを罹患率・発症率回帰ステップと呼ぶ）。次に、Ｒ（ｔ）を用いて（数２）から健康寿命Ｋを算出し（Ｓ１６０）、健康状態を表す抽出条件と健康寿命Ｋを健康状態別の発症リスク知識に書きこむ（Ｓ１７０）。設定Ｓ１２０からＳ１７０までの一連の処理を終えると、他の抽出条件を設定する必要があるかを判定し（Ｓ１８０）、必要がなくなるまで、例えば、“２０（３０、４０）歳代の男性で血糖値が９０未満”、“２０（３０、４０）歳代の男性で血糖値が９０以上１００未満”、“２０（３０、４０）歳代の男性で血糖値が１１０以上１２０未満”のデータを処理し終わるまでＳ１２０からＳ１７０までの処理を繰り返し、処理を終了する（Ｓ１９０）。
【００２６】
図３は、本発明の実施例で用いる回帰式をパラメータの値を３種類設定して（数１）をプロットした図である。図３（ａ）は、パラメータの値がＡ＝０．０５、Ｂ＝０．０５、Ｍ＝３の場合に得られる（数１）の曲線３１０と、パラメータの値がＡ＝０．０５、Ｂ＝０．２、Ｍ＝３の場合に得られる（数１）の曲線３２０とを比較表示し、図３（ｂ）は、曲線３１０と、パラメータの値がＡ＝０．２、Ｂ＝０．０５、Ｍ＝３の場合に得られる（数１）の曲線３３０とを比較表示し、図３（ｃ）は、曲線３１０と、パラメータの値がＡ＝０．０５、Ｂ＝０．０５、Ｍ＝１０の場合に得られる（数１）の曲線３４０とを比較表示している。
【００２７】
実施例１では、発病に関する知見を定式化した回帰式を用いることで、全年齢での罹患率がデータから求められない場合でも、全年齢での罹患率を求めることを可能にする。さらに、実施例１で使用する回帰式は、中年、老年など年齢区分に対応する複数の項を有することで、図３（ａ）の曲線３２０に示すように中年で罹患率が高い場合や、図３（ｂ）の曲線３３０に示すように老年で罹患率が高い場合を表現でき、（数１）のパラメータＭのように、各項のべき数にパラメータを含むことで、図３（ｃ）の曲線３４０に示すように罹患率が急激に増加する場合を表現できるなど、罹患率の様々な変化を表現できる。よって、本発明の処理によれば、少数のデータからも全年齢での罹患率を正確に求めることができる。
【００２８】
実施例１によれば、発症に関する知見を定式化した回帰式を使用することにより、少数のデータからも正確に罹患率又は発症率を算出可能になり、地域別あるいは健康状態別に分類して少数になったデータに対しても正確な罹患率あるいは発症率を算出可能になるので、健康状態や地域性と発症リスクの詳細で正確な関係を得ることができる健康状態別の発症リスク知識構築方法を提供できるという効果が得られる。
（実施例２）
図４は、本発明の実施例２の健康管理装置における機能構成の一例を表すブロック図である。
【００２９】
実施例２では、健康管理装置は１個のサーバによって構成され、健康管理装置の運用者が操作して健康状態別の発症リスク知識を構築し、健康管理装置の利用者が操作して本人の健康状態に対する発症リスクを表示させると仮定する。
【００３０】
図４において、４１０は、運用者と利用者がデータ入力を行うためのフロッピー（登録商標）ドライブ、キーボードなどの入力手段であり、運用者が蓄積健康・罹患データをデータベースに入力するための蓄積用健康・罹患データ入力手段４１１（後述する）と、自分の健康状態に対する発症リスクを知るために利用者が自分の健康・罹患データを検索手段に入力するための参照用健康・罹患データ入力手段４１２（後述する）から成る。４２０は、運用者の入力した健康・罹患データを蓄積するためのデータベースである蓄積健康・罹患データＤＢである。４３０は、実施例１で説明した発症リスク知識構築方法により健康状態別の発症リスク知識を構築するためのプログラムなどである健康状態別の発症リスク知識構築手段である。４４０は、健康状態別の発症リスク知識構築手段により構築した健康状態別の発症リスク知識をテーブル形式で主記憶装置に記憶したファイルなどである健康状態別の発症リスク知識である。４５０は、利用者が入力した健康・罹患データが示す健康状態に対応する発症リスクを得るために健康状態別の発症リスク知識を検索するためのプログラムなどである検索手段である。４６０は、検索手段が獲得した発症リスクを表示するためのモニタなどの表示手段である。
【００３１】
蓄積用健康・罹患データ入力手段４１１は、蓄積健康・罹患データＤＢ４２０に蓄積したい健康・罹患データを蓄積健康・罹患データＤＢ４２０へ入力する。蓄積健康・罹患データＤＢ４２０は、蓄積用健康・罹患データ入力手段４１１から入力された健康・罹患データを蓄積健康・罹患データとして蓄積する。発症リスク知識構築手段４３０は、実施例１で説明した発症リスク知識構築方法により、蓄積健康・罹患データＤＢ４２０から一定の健康状態にあるデータを抽出して発症リスクを算出し、健康状態別の発症リスク知識４４０に書き込む。健康状態別の発症リスク知識４４０は、発症リスク知識構築手段４３０が算出した健康状態別の発症リスク知識を保持する。参照用健康・罹患データ入力手段４１２は、自分の健康状態に対する発症リスクを知りたい利用者の健康・罹患データを検索手段４５０へ入力する。検索手段４５０は、参照用健康・罹患データ入力手段４１２から入力された健康・罹患データの示す健康状態を判断し、健康状態別の発症リスク知識４４０を検索して、判断した健康状態に対応する発症リスクを読み出す。表示手段４６０は、検索手段４５０が健康状態別の発症リスク知識４４０から読み出した発症リスクを表示する。
【００３２】
実施例２での処理の一例を説明する。まず、健康管理装置の運用者の行う処理を説明する。運用者は随時、例えば、健康診断ごとに蓄積用健康・罹患データ入力手段を用いて蓄積健康・罹患データＤＢ４２０に健康・罹患データを入力し、蓄積健康・罹患データＤＢ４２０を更新する。さらに、定期的、例えば、１年に１回、健康状態別の発症リスク知識構築手段４３０を起動して蓄積健康・罹患データＤＢ４２０から一定の健康状態にあるデータを抽出して発症リスクを算出し、健康状態健康状態別の発症リスク知識４４０を作成又は更新する。
【００３３】
次に、健康管理装置の利用者の行う処理を説明する。利用者は自分の健康状態に対する発症リスクを知りたい時、例えば、健康診断時に、参照用健康・罹患データ入力手段４１２から自分の健康・罹患データを検索手段４５０へ入力する。すると検索手段４５０が参照用健康・罹患データ入力手段４１２から入力された健康・罹患データの示す健康状態を判断し、健康状態別の発症リスク知識４４０を検索して、判断した健康状態に対応する発症リスクを読み出す。これを受けて表示手段４６０が検索手段４５０の読み出した発症リスクを表示する。
【００３４】
実施例２によれば、実施例１で説明した発症リスク知識構築方法を用いて構築した健康状態別の発症リスク知識を活用することで、健康状態や地域性と発症リスクの詳細で正確な関係に基づいて健康管理に寄与する健康管理支援装置を提供できるという効果が得られる。
【００３５】
【発明の効果】
本発明によれば、発症に関する知見を定式化した回帰式の使用により少数のデータからも正確に罹患率又は発症率を算出可能になり、地域別あるいは健康状態別に分類して少数になったデータに対しても正確な罹患率あるいは発症率を算出可能になるので、健康状態や地域性と発症リスクの詳細で正確な関係を得ることができる、健康状態別の発症リスク知識構築方法及び健康状態別の発症リスク知識を活用して健康管理に寄与する健康管理支援装置を提供できる。
【図面の簡単な説明】
【図１】本発明の実施例１の処理フローの一例を説明する図。
【図２】本発明の実施例１で使用される蓄積データの一例を示す図。
【図３】本発明の実施例１で用いる回帰式をパラメータの値を３種類設定してプロットした図。
【図４】本発明の実施例２の健康管理装置における機能構成の一例を表すブロック図。
【符号の説明】
２１０…個人の属性を表す情報、２１１…個人特定ＩＤ、２１２…検査特定ＩＤ、２２０…検査属性、２２２…抽出条件特定ＩＤ、２３０…検査値の情報、２４０…問診の情報、２５０…現在の罹患状態の情報、２６０…個人の属性を表す項目についての命題、２７０…検査値の情報の項目についての命題、２８０…問診の情報の項目についての命題、２９０…健康状態を表す情報、２９５…健康寿命、４１０…入力手段、４１１…蓄積用健康・罹患データ入力手段、４１２…参照用健康・罹患データ入力手段、４２０…蓄積健康・罹患データＤＢ、４３０…健康状態別の発症リスク知識構築手段、４４０…健康状態別の発症リスク知識、４５０…検索手段、４６０…表示手段。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an onset risk knowledge construction method for each health condition that builds correspondence knowledge between the health condition and the risk of developing lifestyle-related diseases, and a health management device that contributes to health management by utilizing the onset risk knowledge for each health condition.
[0002]
[Prior art]
In a health management device, it is important to consider individual characteristics in order to motivate health management. Conventionally, there is a report on a method for constructing onset risk knowledge for each health state and a health management device that contribute to health management using knowledge of correspondence between the health state, the health degree, and the onset risk (for example, Patent Document 1 and Patent Document 2). In Patent Document 1, a disease model and a standard model are created from a large number of data, and the health level of the individual is evaluated by comparing the two models with the individual medical examination data. In Patent Document 2, health data is converted into risk of onset using certain rules such as the results of large-scale drug intervention tests and equations derived from epidemiological survey results.
[0003]
[Patent Document 1]
JP 2002-63278 A [Patent Document 2]
Japanese Patent Laid-Open No. 2002-24401
[Problems to be solved by the invention]
According to the above-described conventional technology, it is possible to provide a health degree or onset risk in consideration of individual characteristics. However, because the risk of onset cannot be calculated accurately from a small number of data, the cause of the difference between the health data and the risk of developing the disease from the national average, such as the locality of the user who is provided with the health or the risk of onset, is known. There were two problems: it was difficult to reflect the factors, and the problem that it was difficult to accurately determine the difference in risk of onset for slight differences in health status. The present invention has been made in consideration of such circumstances.
[0005]
The purpose of the present invention is to reflect factors that cause a difference from the national average such as the regional characteristics of users, and further to accurately determine the difference in risk of onset for even slight differences in health status. It is in providing the onset risk knowledge construction method and health management apparatus.
[0006]
[Means for Solving the Problems]
(1) The onset risk knowledge construction method for each health condition according to the present invention has the following characteristics in the onset risk knowledge construction method for each health condition that builds correspondence knowledge between the health condition and the onset risk of lifestyle-related diseases. ing. Personal basic information, which is basic information that is related to the onset risk such as gender and date of birth, and lifelong invariant information, and test results that are the result of quantitatively measuring the current health condition such as blood pressure and weight using equipment etc., Using the interview results, which are the results of collecting current lifestyle habits and health conditions, such as meal content and stress, using questionnaires and interviews, and the presence / absence of disease and disease information Analyzing the relationship between basic information, test results, interview results, and morbidity information for each individual, and specifically calculating the morbidity rate or incidence rate using a regression formula that formulates knowledge about the onset -Having an incidence regression step.
[0007]
(2) The health management device of the present invention has the following features. In the health management device that contributes to health management by utilizing knowledge of the correspondence between the health status and the risk of developing lifestyle-related diseases, basic personal information that is basic information that is invariant throughout life related to the onset risk such as gender and date of birth, Test results that are the results of measuring current health conditions such as blood pressure and weight using equipment, etc., and results of collecting current lifestyle and health conditions such as meal contents and stress using questionnaires and interviews The morbidity / onset rate regression unit calculates the relationship between basic personal information, test results, interrogation results, and morbidity or incidence, using the interrogation results and the presence / absence of the current disease have. The morbidity / onset rate regression unit analyzes the relationship between individual basic information, test results, interview results, and morbidity information for each individual, specifically, using a regression formula that formulates knowledge about onset, Calculate morbidity or incidence.
[0008]
The above (1) and (2) have the following characteristics.
(B) The regression equation has a plurality of terms corresponding to age categories such as middle age and old age, and the power of each term is a parameter or a constant determined based on data.
(B) When L (t) is the survival rate at age t, α1 and α2 are upper limit values of survivable ages, such as 100, and A, B, and M are parameters, the regression equation is ) And represents the morbidity R (t) at age t.
[0009]
[Expression 1]

[0010]
By using regression formulas that formulate knowledge about the onset, it is possible to accurately calculate the morbidity rate or incidence rate from a small number of data, and it is accurate even for data that has become a small number classified by region or health status It is possible to calculate a precise morbidity rate or onset rate, and to obtain a detailed and accurate relationship between the health status, regional characteristics, and onset risk. As a result, it is possible to provide a health management support apparatus that contributes to health management by utilizing the onset risk knowledge construction method for each health condition and the onset risk knowledge for each health condition.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
The onset risk knowledge construction method by health status of the present invention is based on the basic personal information that is related to the onset risk such as gender and date of birth, and the current health status such as blood pressure and weight, etc. The results of quantitative measurements using the test results, the results of interviews that are the results of collecting current lifestyle and health conditions such as dietary content and stress using questionnaires and interviews, and current diseases From the accumulated health / morbidity data that accumulated a lot of health / morbidity data including the presence / absence of the disease and the morbidity information that is the name of the disease, we analyzed the relationship between the individual basic information and test results and the inquiry results and morbidity information by individual, This is a method of constructing knowledge of onset risk by health condition, such as morbidity, to calculate the onset risk, and the morbidity / onset rate regression step to calculate the morbidity or incidence by using a regression formula that formulates knowledge about onset. Having.
[0012]
The health management device of the present invention is a health management device that contributes to health management by utilizing the correspondence knowledge between the health condition and the onset risk of lifestyle-related diseases, and uses the onset risk knowledge construction method described above. The data used for constructing the onset risk knowledge is the health / morbidity data accumulated in the facility where the device is operated or the data set by the operator of the device, and the health constructed by the onset risk knowledge construction method using these data Use the knowledge of the onset risk by condition as the onset risk knowledge by health status unique to the facility. When the personal health / morbidity data is input, the apparatus of the present invention searches for onset risk knowledge for each health condition, and acquires and outputs the onset risk for the input health / morbidity data. The input data of the health management device of the present invention may be health data that is the same data as the health / morbidity data except that it does not include current disease information.
[0013]
Hereinafter, an example of the onset risk knowledge construction method of the present invention will be described in detail with reference to the drawings.
(Example 1): Description of the onset risk knowledge construction method for each health condition.
[0014]
In the following description, for ease of explanation, the onset risk is represented by the healthy life span H (t) of diabetes. The healthy life span of diabetes is the average number of years until the onset of diabetes, and is an amount calculated from (Equation 2) from the prevalence of diabetes.
[0015]
[Expression 2]

[0016]
(Equation 2), t is age, K (t) is healthy life expectancy at age t, L (t ′) is survival rate at age t ′, R (t ′) is morbidity at age t ′ ((Equation 1 ), And α is the upper limit of viable age, such as 100, for example.
[0017]
FIG. 2 is a diagram showing an example of accumulated data used in the onset risk knowledge construction method according to the embodiment of the present invention.
[0018]
FIG. 2 (a) shows accumulated health / morbidity data representing basic personal information, health status and morbidity information. Basic personal information (information indicating individual attributes) 210 such as gender and age, examination date and examination Examination basic information (examination attributes) 220 such as facilities, examination results (examination information) 230 such as height, blood pressure, physiological examination values, interview results (interview information) 240 such as lifestyle habits, past histories, and family histories , Including current disease state information 250. Accumulated health / morbidity data is managed by adding, for example, “N0000001” as an identification number (212), which is a serial number, and “P0000001”, for example, as an individual identification ID (211). Can be determined uniquely. The relationship between the individual identification ID and the inspection identification ID is one-to-many correspondence. That is, there are two data with the inspection identification IDs “N0000001” and “N0000002” among the data with the individual identification ID “P0000001”. This one-to-many correspondence is a result of accumulating the results of one person undergoing multiple tests.
[0019]
FIG. 2 (b) is an extraction condition for classifying the accumulated health / morbidity data by health state or for extracting the health / morbidity data in the health state of interest. For example, the basic information of the individual, the examination result, the inquiry result, and the

propositions

260, 270, and 280 for each item are created. A proposition 260 indicates a proposition regarding an item representing an individual attribute, a proposition 270 indicates a proposition regarding an examination value information item, and a proposition 280 indicates a proposition regarding an inquiry information item. As an extraction condition, for example, a condition in which the sex is male, the age is 20 to 30 years old, the pulse is 50 or less, the blood glucose level is 90 or less, and the exercise amount per week is 2000 kcal or more is created as an example For example, “C0000001” is added and managed as the extraction condition identification ID (222) which is a serial number.
[0020]
FIG. 2C shows the relationship between the health condition and the onset risk, that is, the onset risk knowledge for each health condition that holds the correspondence between the extraction condition and the healthy life span, and information 290 representing the health condition, for example, management of the extraction condition Number and healthy life expectancy 295. For example, in accordance with the condition C0000001, a lot of data such as data with an examination specific ID “N0000003” and data with an examination specific ID “N0000004” are extracted, and from the extracted data, the column of morbidity information / diabetes Count data of “under treatment” and data of “health” to obtain age-specific morbidity Ri (25) at age ti = 25, and Ri (35), Ri (45) similarly obtained according to conditions C0000002 and C00000003 ) Are regressed by (Equation 1) to obtain, for example, parameters A = 0.3, B = 0.1, M = 2.2, and then healthy life expectancy by (Equation 2) As a result of calculation, “55”, “68”, and “79” are obtained. Further, similarly, the extraction condition “C0000004” or later representing another health condition is set, and the healthy life span is obtained as “51”.
[0021]
[Equation 3]

[0022]
Here, the age t and the morbidity rate R in (Equation 1) and (Equation 2) have no subscript, and the age ti and the morbidity rate Ri in (Equation 3) have the subscript i. This is because the age has a value, whereas the latter has a value corresponding to the number N of data used for regression. ti is an intermediate value between the maximum age and the minimum age of the data used on the right side of (Equation 3), that is, a representative value of the age class. The age-specific survival rate L (t) used in (Equation 1) and (Equation 2) may be a mortality rate of a group providing a healthy life expectancy calculated by some means, or an average value, for example, It may be a value calculated from the life table of the Ministry of Health, Labor and Welfare.
[0023]
FIG. 1 is a diagram illustrating an example of a processing flow of the onset risk knowledge construction method according to the embodiment of the present invention.
[0024]
When the process is started (S110), the extraction condition for extracting the accumulated health / morbidity data in the health state of interest by defining the health state of interest, such as “a male in his twenties, whose blood glucose level is less than 90” And the accumulated health / morbidity data satisfying this extraction condition is extracted (S120). Next, the age-specific morbidity rate Ri (ti) with respect to the age ti is calculated by (Equation 3) (S130). Next, it is determined whether the number of points Ri (ti) used for the regression analysis is sufficient (S140), and S120 and S130 are repeated until it is sufficient. This loop is provided because several points are generally required to perform regression. For example, “blood glucose level is less than 90 for men in their 20s”, “blood glucose for men in their 30s” The morbidity rates R (25), R (35), and R (45) by age corresponding to “value is less than 90” and “blood glucose level is less than 90 for men in their 40s” are calculated.
[0025]
When the determination S140 is passed, parameters A, B, and M are obtained from Ri (ti) by regression analysis that represents the morbidity according to (Equation 1), and the morbidity R (t) at all ages can be calculated. The morbidity rate R (t) necessary to calculate K is calculated (S150, this step is called the morbidity / onset rate regression step). Next, the healthy life span K is calculated from (Equation 2) using R (t) (S160), and the extraction condition representing the health state and the healthy life span K are written in the onset risk knowledge for each health state (S170). When a series of processing from setting S120 to S170 is finished, it is determined whether other extraction conditions need to be set (S180). For example, a male in the age group “20 (30, 40)” Blood glucose level is less than 90 "," 20 (30, 40) men in their 20s, blood glucose level is 90 or more and less than 100 "," 20 (30, 40) men in their 20s, blood glucose level is 110 or more and less than 120 " Until the process is completed, the processes from S120 to S170 are repeated, and the process is terminated (S190).
[0026]
FIG. 3 is a diagram in which (Equation 1) is plotted by setting three types of parameter values for the regression equation used in the embodiment of the present invention. FIG. 3A shows a curve 310 (Equation 1) obtained when the parameter values are A = 0.05, B = 0.05, and M = 3, and the parameter values are A = 0.05, The curve 320 of (Equation 1) obtained in the case of B = 0.2 and M = 3 is compared and displayed. FIG. 3B shows the curve 310 and the parameter value A = 0.2, B = 0.05 and M = 3 (Equation 1) obtained by comparison with the curve 330 are displayed. FIG. 3C shows the curve 310 and parameter values A = 0.05, B = 0. 05, the curve 340 of (Equation 1) obtained when M = 10 is compared and displayed.
[0027]
In Example 1, by using a regression formula that formulates knowledge related to pathogenesis, it is possible to obtain the morbidity rate at all ages even when the morbidity rate at all ages cannot be obtained from the data. Furthermore, when the regression equation used in Example 1 has a plurality of terms corresponding to age categories such as middle age and old age, the morbidity rate is high in middle age as shown by the curve 320 in FIG. Or, as shown by a curve 330 in FIG. 3 (b), a case where the morbidity rate is high due to old age can be expressed, and by including a parameter in the power of each term like the parameter M in (Expression 1), FIG. Various changes in the morbidity rate can be expressed, such as a case where the morbidity rate increases rapidly as shown by the curve 340 in (c). Therefore, according to the process of the present invention, the morbidity rate at all ages can be accurately obtained from a small number of data.
[0028]
According to Example 1, it is possible to accurately calculate the morbidity rate or the incidence rate from a small number of data by using the regression formula that formulates the knowledge about the onset, and categorize it by region or health condition, Since it is possible to calculate an accurate morbidity rate or incidence rate even for data that has become, it is possible to obtain an accurate relationship between the health status and regional characteristics and the risk of onset, and a method for constructing risk knowledge by health status Can be provided.
(Example 2)
FIG. 4 is a block diagram illustrating an example of a functional configuration in the health management apparatus according to the second embodiment of the present invention.
[0029]
In the second embodiment, the health management device is configured by a single server, which is operated by the health management device operator to build onset risk knowledge for each health condition, and the health management device user operates the personal information. Suppose you want to display the risk of developing a health condition.
[0030]
In FIG. 4, reference numeral 410 denotes input means such as a floppy (registered trademark) drive and a keyboard for the operator and the user to input data, and storage for the operator to input stored health / morbidity data to the database. Health / morbidity data input means 411 (to be described later) and reference health / morbidity data input means for the user to input his / her health / morbidity data to the retrieval means in order to know the risk of developing his / her health condition 412 (described later). An accumulated health / morbidity data DB 420 is a database for accumulating health / morbidity data input by the operator. 430 is an onset risk knowledge construction means for each health condition, such as a program for constructing onset risk knowledge for each health condition by the onset risk knowledge construction method described in the first embodiment. 440 is onset risk knowledge for each health condition, such as a file in which the onset risk knowledge for each health condition constructed by the onset risk knowledge constructing means for each health condition is stored in the main storage device in a table format. Reference numeral 450 denotes search means such as a program for searching for onset risk knowledge for each health condition in order to obtain an onset risk corresponding to the health condition indicated by the health / morbidity data input by the user. Reference numeral 460 denotes display means such as a monitor for displaying the onset risk acquired by the search means.
[0031]
The health / morbidity data input means 411 for accumulation inputs the health / morbidity data to be accumulated in the accumulated health / morbidity data DB 420 to the accumulated health / morbidity data DB 420. The accumulated health / morbidity data DB 420 accumulates the health / morbidity data input from the accumulation health / morbidity data input means 411 as accumulated health / morbidity data. The onset risk knowledge constructing means 430 calculates onset risk by extracting data in a certain health state from the accumulated health / morbidity data DB 420 by the onset risk knowledge constructing method described in the first embodiment, and the onset by health state. Write to risk knowledge 440. The onset risk knowledge 440 for each health condition holds the onset risk knowledge for each health condition calculated by the onset risk knowledge construction unit 430. The reference health / morbidity data input unit 412 inputs the health / morbidity data of the user who wants to know the onset risk for his / her health condition to the search unit 450. The search unit 450 determines the health state indicated by the health / morbidity data input from the reference health / morbidity data input unit 412, searches the onset risk knowledge 440 for each health state, and corresponds to the determined health state. Read outset risk. The display means 460 displays the onset risk read by the search means 450 from the onset risk knowledge 440 for each health condition.
[0032]
An example of processing in the second embodiment will be described. First, the process performed by the operator of the health management apparatus will be described. The operator inputs health / morbidity data to the accumulated health / morbidity data DB 420 using the accumulation health / morbidity data input means for each health examination, for example, and updates the accumulated health / morbidity data DB 420 at any time. Further, onset regularly, for example, once a year, the onset risk knowledge construction means 430 for each health condition is activated, data on a certain health condition is extracted from the accumulated health / morbidity data DB 420, and the risk of occurrence is calculated. Then, the onset risk knowledge 440 for each health condition is created or updated.
[0033]
Next, processing performed by the user of the health management device will be described. The user inputs his / her health / morbidity data from the reference health / morbidity data input means 412 to the search means 450 when he / she wants to know the onset risk with respect to his / her health condition, for example, during a health examination. Then, the search unit 450 determines the health state indicated by the health / morbidity data input from the reference health / morbidity data input unit 412, searches the onset risk knowledge 440 for each health state, and corresponds to the determined health state. Read outset risk. In response to this, the display unit 460 displays the onset risk read by the search unit 450.
[0034]
According to Example 2, by using the onset risk knowledge for each health condition constructed using the onset risk knowledge construction method described in Example 1, detailed and accurate relationship between the health condition, regional characteristics, and onset risk Based on the above, it is possible to provide a health management support device that contributes to health management.
[0035]
【The invention's effect】
According to the present invention, it becomes possible to accurately calculate the morbidity rate or the incidence rate from a small number of data by using a regression formula that formulates knowledge about the onset, and the data that has been reduced to a small number by classification by region or health condition Since it is possible to calculate an accurate morbidity rate or incidence rate, it is possible to obtain an accurate relationship between the health status, regional characteristics, and risk of onset, and a method for constructing risk knowledge by health status and health status It is possible to provide a health management support device that contributes to health management by utilizing another onset risk knowledge.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating an example of a processing flow according to a first embodiment of the present invention.
FIG. 2 is a diagram showing an example of accumulated data used in Embodiment 1 of the present invention.
FIG. 3 is a graph obtained by plotting the regression equation used in Example 1 of the present invention by setting three types of parameter values.
FIG. 4 is a block diagram illustrating an example of a functional configuration in the health management apparatus according to the second embodiment of the present invention.
[Explanation of symbols]
210 ... Information representing individual attributes, 211 ... Individual identification ID, 212 ... Examination identification ID, 220 ... Examination attribute, 222 ... Extraction condition identification ID, 230 ... Examination value information, 240 ... Interrogation information, 250 ... Current Information on disease state, 260 ... Proposition about item representing individual attribute, 270 ... Proposition about item of test value information, 280 ... Proposition about item of inquiry information, 290 ... Information representing health condition, 295 ... Healthy life, 410 ... Input means, 411 ... Health / morbidity data input means for accumulation, 412 ... Health / morbidity data input means for reference, 420 ... Accumulation health / morbidity data DB, 430 ... Development risk knowledge construction means by health condition 440... Onset risk knowledge by health state, 450... Search means, 460.

Claims

In the onset risk knowledge construction method by health condition that builds correspondence knowledge between health condition and risk of developing lifestyle-related diseases, basic personal information that is basic information that is invariant throughout life related to onset risk such as gender and date of birth, Test results, which are the results of quantitative measurement of current health conditions such as blood pressure and weight using equipment etc., and current lifestyle and health conditions such as meal contents and stress using questionnaires and interviews Analyzing the relationship between the basic personal information, the test result, the interrogation result, and the morbidity information for each individual using the collected interrogation results and the presence / absence of the current disease and the morbidity information that is the disease name The relationship between the basic individual information, the test result, the interrogation result, and the morbidity or incidence is calculated using a regression formula that determines the parameter values and formulates the knowledge about the onset. Health-specific risk knowledge construction method characterized in that it comprises a rate-incidence regression step.

2. The method for constructing knowledge of onset risk according to health status according to claim 1, wherein the regression equation has a plurality of terms corresponding to age categories such as middle age and old age, and the number corresponding to each of the plurality of terms. Alternatively, at least one of the coefficients is a parameter determined based on the morbidity rate or the onset rate calculated from the individual basic information, the test result, the inquiry result, and the morbidity information. How to build knowledge of the risk of onset.

3. The method for constructing knowledge of risk of onset according to health condition according to claim 2, wherein the regression equation is such that L (t) is a survival rate at age t, α1 and α2 are upper limit values for alive age, A, B, When M is a parameter,

A method for constructing knowledge of onset risk according to health condition, wherein the formula represents the morbidity rate R (t) at age t by (Equation 1).

In the health management device that contributes to health management by utilizing knowledge of the correspondence between the health status and the risk of developing lifestyle-related diseases, basic personal information that is basic information that is invariant throughout life related to the onset risk such as gender and date of birth, Test results that are the results of measuring current health conditions such as blood pressure and weight using equipment, etc., and results of collecting current lifestyle and health conditions such as meal contents and stress using questionnaires and interviews And calculating the relationship between the basic personal information, the test result, the interview result, and the morbidity or incidence, using the interrogation result and the presence / absence of the current disease and the disease name. A health management device comprising a rate regression unit.