Design and data

This study was a cross-sectional analysis using RN staffing data, patient data, and hospital data from 2011 and 2012 that forecasted the supply and demand of nurses during Hurricane Sandy. The RN staffing data for each New Jersey hospital were obtained from the New Jersey Department of Health (NJDOH). The NJDOH collects daily ratio of patients per RNs from all hospitals by unit type, which are further aggregated, averaged, and reported at the quarter level. These quarterly patient to RN ratios were used for this analysis.

Patient data were obtained from the State Inpatient Databases (SID). The SID was developed for the Healthcare Costs and Utilization Project and sponsored by the Agency for Healthcare Research and Quality. Although this dataset includes numerous data on patients’ demographics, diagnoses, procedures, and so on, only the number of inpatient discharges from New Jersey hospitals were used in this study.

Hospital data were obtained from the American Hospital Association (AHA) Annual Survey. As a census survey of the U.S. hospitals, the AHA Annual Survey collects comprehensive characteristics of hospitals, including but not limited to staffing, ownership, beds, geographic characteristics, and accreditations and so on.

Sample

Four out of the 72 acute care hospitals in New Jersey without data on RN staffing were excluded. Another three hospitals were reported as one system. Thus, 66 New Jersey hospitals were included for further analysis. Data on all adult and pediatric patients who were discharged from one of the study hospitals during November 2011 and November 2012 were included in the analyses.

Variables and measurements

Hospital characteristics

The hospital characteristics obtained from the 2012 AHA Annual Survey included bed size, high-technology status, teaching status, safety-net hospital, and Magnet designated status. The bed size was a categorical variable, with the categories of <=100 beds, 101–250 beds, and > 250 beds. The high-technology status was a binary variable (yes or no) and defined as whether a hospital implemented electronic health records or not. Teaching status was also a binary variable (yes or no) and was defined as if a hospital was approved to participate in residency and/or internship training by the Accreditation Council for Graduate Medical Education, had a medical school affiliation through the American Medical Association, was a member of the Council of Teaching Hospitals of the Association of American Medical Colleges, had an internship approved by American Osteopathic Association, or had an residency approved by American Osteopathic Association. The safety-net status was a binary variable (yes or no), and a hospital was defined as a safety-net hospital if it provides uncompensated care for the uninsured and most vulnerable patients [18]. The Magnet designated status was a binary variable (yes or no); a hospital is accredited as a Magnet hospital for good nursing care based on a list obtained from the American Nurses Credentialing Center.

Staffing

The average patient to RN ratio of the fourth quarter (October, November, and December) in 2011 and 2012 were used for this analysis. These data were converted to estimate the RN full-time equivalents (FTEs) in November for 2011 and 2012. All hospital units that provide adult and pediatric acute care were categorized into two types: intensive care unit (ICU) and non-ICU. The ICUs included adult ICU, neonatal ICU, and pediatric ICU; and non-ICU included adult open psychiatric unit, adult closed psychiatric, child/adolescent closed psychiatric, intermediate, medical surgical, neonatal-intermediate, normal newborn nursery, obstetrics, pediatrics, and substance abuse.

Patient characteristics

Data on patient characteristics were derived from the SID and included admission unit, length of stay, and a discharge date that occurred in November 2011 or November 2012. Patients were categorized into two groups based on unit type — ICU and non-ICU. The length of stay and the number of discharges were aggregated to the hospital level where the average length of stay and the total number of discharges for each hospital were computed and used in the analyses.

Storm severity of hurricane Sandy

The counties of New Jersey were categorized as moderate, high, and very high impacted areas based on the number of people exposed to surge, the dollars in wind damage, and the volume of rain [19] (Table 1).

Table 1 A list of counties based on the storm severity of hurricane sandy

Analysis

Data from the SID, AHA, and NJDOH were merged to address the purpose of the study. Our models of forecasting were based on previous work that examined the demand for nurses [20]. Specifically, three models were developed to estimate if the supply of nurses in New Jersey hospitals was able to the meet the demand during Hurricane Sandy.

Model 1

We first calculated the observed RN FTEs in November 2012 using the following formula [21]:

[1] Observed RN FTEs Nov. 2012= ( frac{frac{1}{the number of patient s per RN in Quarter 4 of 2012}ast the total patient days in Nov.2012ast 24}{40ast 4ast 0.85} )

Where the number of patients per RN in the 4th quarter of 2012; the total patient days in November 2012 were calculated by multiplying the average length of stay by the total number of discharges in each hospital in November 2012. The denominator of this formula provided the total observed nursing hours in November 2012. By dividing these nursing hours by 40 h per week, 4 weeks per month, and an assumption of 0.85 productive hours, we obtained the observed RN FTEs for each hospital in November 2012.

In an effort to determine if the observed number of RN FTEs were able to meet the demand for nurses, we first estimated the expected RN FTEs in November 2012 by using the observed RN FTEs of November 2011 as baseline using the following formula:

[2] Expected RN FTEs in Nov.2012= ( frac{Observed RN FTEs in Nov.2011}{Total patient days in Nov.2011} ) * Total patient days in Nov. 2012

Where the observed RN FTEs in November 2011 were calculated using formula [1]; the total patient days in November 2011 were calculated by multiplying the average length of stay by the total number of discharges in each hospital in November 2011. The results of dividing the observed RN FTEs in November 2011 by the total patient days in November 2011 provided the observed RN FTEs per patient day in November 2011. Assuming this rate was not changed in November 2012, the expected RN FTEs in November 2012 were estimated by multiplying the rate by the total patient days in November 2012.

To determine whether a hospital had a shortage of RN FTEs, the expected RN FTEs in November 2012 were subtracted from the Observed RN FTEs in November 2012:

Difference 1 = Observed RN FTEs in November 2012 – Expected RN FTEs in November 2012

A negative difference indicating that the observed RN FTEs was smaller than the expected RN FTEs in November 2012, suggesting of a shortage of nurses.

Model 2

Using November 2012 data on the hospitals with a shortage of RN FTEs, we estimated how an increase in the observed RN FTEs could meet hospital demand. This was calculated through a 10% increase of the observed RN FTEs in November 2012:

Difference 2 = Observed RN FTEs in November 2012*110% – Expected RN FTEs in November 2012

Model 3

Similar to Model 2, using November 2012 data from hospitals with a shortage of RN FTEs, we estimated how an increase in the observed RN FTEs could meet hospital demand. This model was calculated through a 20% increase in the observed RN FTEs in November 2012:

Difference 3 = Observed RN FTEs in November 2012*120% – Expected RN FTEs in November 2012

These three model estimations were conducted separately by unit type (ICU and non-ICU) to estimate the shortage of RN FTEs. Descriptive analyses were used to summarize the shortage of RN FTEs by county. All analyses were conducted using STATA/MP 15.1 (StataCorp, College Station, TX).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Disclaimer:

This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (https://www.biomedcentral.com/)