To comply with new inpatient immunization measures introduced by the Centers for Medicare and Medicaid Services (CMS) effective January 1, 2012, our health system implemented a nurse‐driven vaccine protocol using native functionality in our inpatient electronic medical record (EMR) to assess for and administer pneumococcal (pneumo) and influenza (flu) vaccines.


A multidisciplinary team developed the protocol based on current‐state vaccine process mapping and Kano analysis. The protocol was implemented as an evidence‐based clinical decision support (CDS) tool in our inpatient EMR (Allscripts, Chicago, IL) to optimize existing admission work flows. Previously, inpatient pneumo and flu vaccines were ordered by default in the admission order sets unless contraindicated or prior vaccination data were documented in our inpatient or outpatient EMR (Epic, Verona, WI). The new process continued to leverage prior vaccination data but had a nursing vaccine assessment selected by default, a vaccine assessment task and form for the nurse to complete, and functionality for the nurse to submit vaccine orders when needed. The provider could also perform the assessment if desired. The vaccine protocol was implemented on August 20, 2012, for pneumo vaccine, with flu vaccine assessment added at the start of the flu season on September 24, 2012. We examined 3 outcomes: (1) nurse vaccine assessment postimplementation, (2) vaccine administration pre‐ and postimplementation, and (3) pre‐ and postimplementation immunization core measures. Preimplementation data were collected for the 60‐day period from January 24 to March 23, 2012, to align with the new CMS immunization measures and related revisions to our provider‐based vaccine assessment, and postimplementation data were collected for the 60‐day period from September 25 to November 23, 2012. Only 1 month of core immunization data were available postimplementation; therefore, we compared the data available in the last month of the last flu season with the first month of this flu season. Proportions were compared using the chi‐square test.


Approximately 10% of vaccine assessments were satisfied at admission via the automated search of historical vaccination records, approximately 90% of inpatients received the default nurse pneumococcus and influenza vaccine assessment orders, and 85% of nurse assessments were completed by patient discharge (Table 1). When compared with preintervention data, the efficiency of the vaccine assessment and administration process was improved in the postintervention period, and as a result our vaccination core measure rates improved (Table 2).


An automated nurse‐driven vaccine protocol linked to an electronic admission order set increased the efficiency and rates of pneumococcus and influenza vaccination across the inpatient units of a multihospital academic health system. This intervention was created in a commonly used commercial EMR and is scalable across institutions with similar systems.

Table 1.Postimplementation Vaccine Assessments

Pneumococcal vaccine
 Assessment automatically satisfied with historical data 11%
 Nursing assessment selected by default 87%
 Nursing assessment satisfied by discharge 85%
Influenza vaccine
 Assessment automatically satisfied with historical data 10%
 Nursing assessment selected by default 88%
 Nursing assessment satisfied by discharge 85%

Table 2.Vaccine Ordering and Administration Pre- and Postimplementation

Pre Post Absolute Change P Value
Pneumococcal vaccine
 Vaccine ordered 3211 (22%) 1375 (10%) 12% Decrease < 0.001
 Ordered vaccines administered 341 (11%) 872 (63%) 52% Increase < 0.001
 CMS core measure satisfied 69% 87% 18% Increase < 0.01
Influenza vaccine
 Vaccine ordered 7494 (51%) 2996 (22%) 29% Decrease < 0.001
 Ordered vaccines administered 463 (6%) 2020 (67%) 61% Increase < 0.001
 CMS core measure satisfied 72% 79% 7% Increase 0.08