Functional Requirements - System
The system must accept a daily scheduled data drop from Notes via the
hospital network, with adequate error trapping to reject corrupted
files, deal with failures of the network/Notes server, and power-supply
or network malfunctions during data transfer.
Emergency disk-based transfer of files for several days of scheduled
drops must be allowed on a single day with no degradation of
functionality.
Full specifications for data input to the system must be revealed, and
the specification must be sufficiently simple to allow ready adaptation
to future changes in the TelePath/Notes system. Any problems with the
data input, or the integrity of the database must be flagged with clear
error-messages, and readily correctable by the user.
New antibiotics, 'firms', wards, species and hospitals can be added
to the TelePath system that ultimately generates the data, and such
changes must be integrated into any system code lists automatically,
after user confirmation.
The data structure and system should be capable of operating with up to
10 years positive specimens (about 500,000 specimens).
Secure hidden logs of all password entries for 1 month, all parameter
changes for 12 months, all data update transactions for 3 months, all
detected anomalies (short report) and all edits of the data must be
available to two master operators only.
Anomaly Detection - The client requires a system that is able to
examine an ASCII data drop from a Lotus Notes database containing a
subset of the PHL data. The prototype system, should be able to process
the data overnight, and the following morning produce a text file
containing any anomalies found in the data. The file produced should
report any anomalies in bacteria levels that - 1) occur in particular
ward, 2) are associated with a particular doctor, or 3) are associated
with a particular firm.
Scalability - The system should be designed in such a way that further
system requirements can be added to the design and implementation of
the system; such as Correlation Detection, Trending, Supervised
Learning and Data Filtering.
Security - The data contained in the database is personal and
confidential. Steps HAVE to be taken ensure patient confidentiality,
and identifying fields will need to be encrypted.
Robustness - The TelePath database in which the data is stored is
subject to change and the system must be able to cope with future
changes to the underlying database.
A Graphical User Interface - The system should have an extendable GUI
to allow ease of use. The system should be capable of running on a
Windows PC running either Win95/Win98/NT4.0.
Time Dependent - The system should be able to operate and complete
its designated tasks over a nightly period (nominally ~ 6 - 8 hours).
Storage Dependent - The data structure and system should be capable of
operating with up to 10 years positive specimens (about 500,000
specimens).
Project Functional Requirements - Extended System
These requirements are those which should be considered in the design
and implementation of the base system, and added to the prototype if
there is time. However, some of the tasks are particularly complex and
are considered only for later addition as part of a research project or
PhD.
Data Integrity - The system should perform checks to ensure that the
data has not been corrupted, is not inconsistent.
Data Extraction Thresholds - Allow the user to set defined thresholds
on particular features in the data to allow greater control over the
data extracted.
Multi-User - The system shall be able to handle different users, who
may have different data interests. Each user can have a different set
of "interest parameters", thus allowing each user to perform data
extraction using varying levels of interest on particular data
features. This includes the facility to allow a user to continue
extracting data from the time they last used the system.
Advanced Anomaly Detection - The system should be able to give a
reasonable prediction of bacterial outbreaks before they occur, and
estimate the urgency of the outbreak.
Correlation Detection, Trending and Supervised Learning - The system
should be able to search for correlations and trends in the data. An
operator could guide the search by changing "threshold levels" for a
particular extraction, and with the knowledge of confirmed outbreaks
train the system to detect certain expected trends in the data, or try
and discover new trends.
System Portability - Abstract the design and implementation of the
system to allow possible porting of the system to another PHL with a
similar system.
医 院 感 染 学 是 近 年 发 展 起 来 的 一 门 重
要 的 新 兴 学 科 , 是 当 代 临 床 医 学 、 预 防 医
学 和 医 院 管 理 学 的 一 个 重 大 课 题 [1]。 医 院
感 染 控 制 的 好 坏 , 将 直 接 影 响 医 院 的 整 体
医 疗 水 平 , 并 关 系 到 医 院 的 发 展 前 景 [2]。
但 目 前 , 医 院 感 染 监 控 多 数 是 回 顾 性 的 ,
对 医 院 感 染 流 行 和 爆 发 的 可 能 及 趋 势 不 能
进 行 前 瞻 性 监 测 , 对 控 制 和 减 少 医 院 感 染
的 感 染 率 起 不 到 应 有 的 作 用 。 为 改 变 目 前
这 种 被 动 局 面 , 我 院 通 过 应 用 " 军 卫 一 号
" 工 程 医 院 信 息 系 统 子 系 统 " 医 院 感 染 监
控 管 理 软 件 " , 对 医 院 感 染 危 险 因 素 进 行
实 时 监 测 和 重 点 监 控 , 实 现 了 感 染 管 理 、
感 染 登 记 、 感 染 相 关 因 素 统 计 等 监 测 项 目
所 需 的 各 项 功 能 [3]。 医 院 感 染 监 控 管 理 软
件 的 应 用 , 不 仅 能 对 医 院 感 染 相 关 因 素 进
行 主 动 、 连 续 、 系 统 地 监 测 , 还 可 为 感 染
控 制 专 职 人 员 提 供 极 大 的 便 利 , 提 高 医 院
感 染 管 理 的 水 平 [4]。
1 在 科 室 管 理 上 的 应 用
1 1 制 订 感 染 质 量 评 估 标 准
依 据 《 医 院 感 染 管 理 规 范 ( 试 行 ) 》
的 要 求 , 该 管 理 软 件 提 供 了 较 为 完 善 的 医
院 感 染 检 查 内 容 和 评 分 标 准 , 各 科 室 可 根
据 情 况 选 择 使 用 , 并 在 此 基 础 上 制 订 各 自
的 检 查 内 容 ; 同 时 , 各 科 室 利 用 现 有 平 台
, 也 可 看 到 对 自 己 的 检 查 项 目 及 评 分 标 准
, 有 助 于 展 开 自 查 , 使 感 染 控 制 工 作 真 正
成 为 人 们 日 常 工 作 的 一 部 分 。
1 2 查 询 环 境 送 检 结 果
查 询 各 科 室 的 空 气 、 器 皿 的 细 菌 培 养
检 测 结 果 , 检 测 结 果 能 自 动 反 映 到 感 染 检
查 记 录 中 。
1 3 感 染 检 查 分 析
记 录 各 科 室 感 染 检 查 结 果 , 结 合 自 身
的 环 境 卫 生 学 监 测 , 可 方 便 对 科 室 感 染 管
理 情 况 进 行 评 估 ; 同 时 , 能 自 动 汇 总 医 院
感 染 检 查 中 各 科 室 存 在 的 普 遍 问 题 , 便 于
展 开 有 目 的 的 培 训 和 教 育 。
2 在 日 常 监 控 工 作 中 的 应 用
2 1 在 院 病 人 查 询
通 过 该 管 理 软 件 的 应 用 , 可 随 时 掌 握
在 院 病 人 的 年 龄 、 病 情 、 入 院 诊 断 、 住 院
天 数 、 最 高 体 温 、 腹 泻 、 血 常 规 检 查 中 的
白 细 胞 计 数 值 等 。 最 高 体 温 的 设 定 , 可 以
是 最 近 3天 或 更 长 时 间 内 的 ( 一 般 设 定 为
3天 ) ; 血 常 规 检 查 是 最 近 2次 的 结 果 , 并
给 出 了 相 应 的 检 查 日 期 。 可 以 住 院 天 数 、
体 温 、 科 室 为 条 件 对 病 人 进 行 筛 选 。 这 为
有 目 的 地 展 开 监 控 工 作 提 供 了 方 便 。
2 2 在 用 抗 生 素 查 询
通 过 该 管 理 软 件 的 应 用 , 可 随 时 掌 握
在 院 病 人 的 抗 菌 药 物 应 用 情 况 , 包 括 病 人
所 在 科 室 、 入 院 诊 断 、 细 菌 鉴 定 结 果 、 药
品 名 称 、 用 药 开 始 时 间 、 剂 量 、 用 药 途 径
、 频 次 、 限 定 日 剂 量 ( Definded daily dose, DDD)
。 可 以 抗 菌 药 物 合 并 用 药 数 、 DDD值 、 用 药
天 数 、 药 品 名 称 为 条 件 对 病 人 进 行 筛 选 ,
便 于 及 时 掌 握 各 科 室 的 抗 菌 药 物 使 用 情 况
, 对 其 合 理 用 药 进 行 指 导 。 例 如 , 要 查 询
谁 在 使 用 " 万 古 霉 素 " , 只 需 在 药 品 名 称
栏 按 " F9" 键 , 输 入 拼 音 字 头 " WGMS" , 就
可 显 示 出 使 用 该 药 的 病 人 的 列 表 。 如 果 对
查 询 结 果 感 兴 趣 , 还 可 双 击 查 看 病 人 在 院
的 病 历 、 治 疗 记 录 、 检 查 报 告 、 检 验 结 果
等 , 从 而 做 到 工 作 有 重 点 、 有 效 果 。
2 3 在 用 介 入 性 操 作 查 询
通 过 该 管 理 软 件 的 应 用 , 可 随 时 掌 握
在 院 病 人 的 介 入 性 操 作 使 用 情 况 , 包 括 病
人 所 在 科 室 、 入 院 诊 断 、 操 作 名 称 、 开 始
时 间 、 频 次 、 操 作 医 师 职 称 、 执 行 护 士 职
称 。 可 以 种 类 数 、 使 用 天 数 、 操 作 名 称 为
条 件 对 病 人 进 行 筛 选 , 掌 握 各 科 室 的 介 入
性 操 作 使 用 情 况 。
3 数 据 统 计 分 析
3 1 数 据 统 计
应 用 该 管 理 软 件 , 可 对 医 院 感 染 的 诸
因 素 进 行 统 计 , 对 临 床 治 疗 有 着 十 分 重 要
的 指 导 意 义 。 它 可 做 到 图 文 并 茂 , 图 表 有
全 院 临 床 科 室 医 院 感 染 总 表 、 全 院 手 术 科
室 医 院 感 染 总 表 、 全 院 临 床 科 室 感 染 部 位
及 构 成 比 统 计 表 、 全 院 临 床 科 室 抗 生 素 使
用 报 表 、 全 院 细 菌 药 敏 检 验 表 、 全 院 细 菌
检 出 分 布 表 及 全 院 感 染 与 特 殊 用 药 等 。 例
如 , 全 院 临 床 科 室 抗 生 素 使 用 报 表 可 统 计
某 一 段 时 期 内 全 院 内 科 、 外 科 、 各 临 床 科
室 各 种 抗 生 素 的 使 用 人 数 ( 其 中 单 列 出 感
染 病 人 使 用 数 ) 、 次 数 , 抗 生 素 的 一 联 、
二 联 、 多 联 使 用 情 况 , 平 均 DDD值 、 使 用 天
数 、 平 均 费 用 等 , 并 可 按 条 件 进 行 排 序 。
这 样 , 也 可 为 临 床 药 师 了 解 医 院 的 抗 菌 药
物 用 药 趋 势 提 供 帮 助 。
3 2 药 敏 数 据 分 析
应 用 该 管 理 软 件 , 可 根 据 送 检 日 期 、
患 者 信 息 、 送 检 科 室 等 汇 总 某 种 ( 类 ) 细
菌 的 药 敏 结 果 。 如 , 可 以 2005年 1月 1日 ~
2005年 6月 30日 为 时 间 段 , 或 男 性 、 >60岁 或
<10岁 、 每 个 病 人 取 最 大 耐 药 结 果 、 革 兰 阳
性 球 菌 但 不 包 括 甲 型 链 球 菌 、 亚 胺 培 南 耐
药 、 超 广 谱 β - 内 酰 胺 酶 阳 性 为 条 件 , 统
计 出 药 敏 结 果 离 散 图 。 利 用 该 项 功 能 , 可
方 便 统 计 出 细 菌 耐 药 趋 势 , 为 指 导 临 床 用
药 提 供 参 考 。
3 3 患 者 资 料 分 析
根 据 出 院 时 间 、 出 院 诊 断 、 出 院 科 室
, 统 计 出 院 病 人 的 抗 菌 药 物 使 用 情 况 、 感
染 发 生 与 季 节 、 年 龄 、 住 院 时 间 、 术 前 占
床 日 、 抗 菌 药 物 、 介 入 操 作 、 原 发 疾 病 、
化 疗 、 放 疗 、 免 疫 抑 制 药 等 的 关 系 , 及 分
析 统 计 撰 写 论 文 的 全 部 资 料 。
3 4 数 据 查 询
根 据 系 统 记 录 的 病 人 资 料 , 如 病 案 首
页 、 抗 菌 药 物 使 用 、 介 入 性 操 作 使 用 、 检
验 结 果 、 细 菌 培 养 、 药 敏 报 告 信 息 等 查 询
病 人 的 相 关 资 料 。 如 , 可 利 用 该 系 统 查 询
出 烧 伤 病 人 、 细 菌 培 养 中 检 出 " 鲍 曼 不 动
杆 菌 " 且 该 菌 对 " 亚 胺 培 南 " 耐 药 的 病 人
的 抗 菌 药 物 使 用 资 料 。 这 些 资 料 可 导 出 成
Excel文 件 , 为 进 一 步 利 用 SPSS或 SAS之 类 的 社 会
化 统 计 分 析 软 件 处 理 数 据 提 供 支 持 , 为 开
展 深 层 次 的 研 究 提 供 方 便 。
4 软 件 应 用 中 须 注 意 的 问 题
4 1 数 据 维 护
必 须 定 期 维 护 系 统 数 据 。 因 为 医 院 的
抗 菌 药 物 品 种 并 非 固 定 不 变 , 新 进 抗 菌 药
物 品 种 必 须 及 时 增 加 到 统 计 项 目 中 , 否 则
就 不 能 收 集 到 病 人 抗 菌 药 物 使 用 库 中 。
4 2 日 期 范 围 定 义
系 统 的 统 计 数 据 包 含 截 止 日 期 发 生 的
数 据 , 在 定 义 日 期 范 围 时 要 注 意 正 确 理 解
, 否 则 容 易 和 其 它 系 统 统 计 的 数 据 出 现 不
一 致 的 情 况 。
4 3 抗 菌 药 物 的 合 理 应 用 判 断
抗 菌 药 物 合 理 应 用 的 质 量 标 准 , 是 以
用 药 品 种 、 用 药 量 、 使 用 经 费 、 不 良 反 应
及 药 物 相 互 作 用 等 进 行 综 合 性 评 定 , 系 统
不 能 实 现 自 动 判 断 , 需 要 人 工 录 入 。
5 结 语
运 用 医 院 感 染 监 控 管 理 软 件 , 可 实 现
医 院 相 关 医 疗 信 息 共 享 。 通 过 采 集 其 中 与
医 院 感 染 相 关 的 实 时 数 据 ( 危 险 因 素 信 息
) , 对 其 进 行 汇 总 、 分 析 , 做 出 预 警 , 切
实 为 临 床 科 室 提 供 预 防 感 染 提 示 , 从 而 为
预 防 和 控 制 医 院 感 染 、 降 低 医 院 感 染 发 生
率 提 供 重 要 依 据 , 提 高 医 院 感 染 管 理 人 员
的 工 作 效 率 和 增 强 效 果 ; 同 时 , 它 还 可 向
医 院 管 理 层 提 供 数 据 接 口 , 成 为 医 院 数 字
化 管 理 的 重 要 组 成 部 分 。 该 软 件 在 我 院
3年 多 的 使 用 过 程 中 , 取 得 了 较 好 的 实 际
效 果 并 积 累 了 大 量 的 数 据 , 为 下 一 步 分 析
医 院 感 染 发 生 的 原 因 、 感 染 途 径 与 防 控 方
法 和 医 院 感 染 管 理 控 制 的 成 本 与 效 益 评 价
、 建 立 耐 药 菌 株 与 抗 生 素 治 疗 的 卫 生 经 济
学 模 型 等 研 究 打 下 了 基 础 [5], 具 有 较 大 的
科 研 潜 能 。
参 考 文 献
[1] 李 六 亿 医 院 感 染 监 测 的 发 展 与 展
望 [J] 中 华 医 院 感 染
学 杂 志 , 1999, 9( 6) : 223
[2] Scheckler WE, Brimhall D, Buck AS, et al Requireme_
nts for infrastructure and essential activities of infection
control and epidemiology in hospitals: a consensure panel report
Society for healthcare epidemiology of America[J] Infect Control
Hosp Epidemiol, 1998, 19( 2) : 114
[3] 吴 超 浅 析 现 阶 段 医 院 感 染 监 控 工
作 中 的 若 干 不 足 [J]
江 苏 卫 生 事 业 管 理 , 2001, 12( 4) : 11
[4] 金 玉 平 信 息 反 馈 在 医 院 感 染 管 理
中 的 作 用 [J] 中 华 医 院 感 染 学 杂 志 , 2002,
12( 4) : 302
[5] 朱 士 俊 医 院 感 染 管 理 的 现 状 与 展
望 [J] 中 国 医 院 ,
2004, 8( 1) : 8