presence of eccentric or concentric stenosis. Using a cell-path knife, the kidney was cut in a frontal plane to allow for inspection of the parenchyma. If lesions were palpable but not visibly seen in the section, additional slices were cut to examine the parenchyma. One tissue sample was taken from kidney, extending from the cortical surface to the minor calyx. Additional tissue samples of lesion areas were also taken during the inspection process. Histologic examination of organs The tissue cassettes, disposable plastic containers used to hold and identify tissue samples, were processed and the tissues were embedded with paraffin wax using a standard overnight protocol using an Excelsior ES tissue processor. After placing the tissue in paraffin blocks using a Shandon Histocentre 3, the tissue was sectioned at five micrometers using a Leica RM 2135 microtome and manually stained with Hematoxylin and Eosin. A board certified pathologist examined the slides for histopathology (two pathologists with 10+ years experience and a senior resident). The clinicians were blinded as to the gross observations that were described with each case. Methods of qualitative analyses All of the gross observations were placed in categories using a qualitative content analysis approach that included both an open coding process followed by a selective coding process (see below). Content analysis is a method of qualitative research where words are coded, grouped and recoded to identify themes that exist within documents. For this process, an anatomist (PhD trained) with experience in qualitative research methodology read the written notes that were compiled after the gross and histologic evaluations. The documents were then coded in a two-phase process as described below. Open coding process Each individual raw observation was taken from the notes made at the time of gross examination of the kidney. The observations were listed in their original language. Then, they were grouped together based on common themes and attributes. For example, individual categories such as “brown discoloration” and “black discoloration” were grouped together into one category, “brown or black discoloration”. There were three rounds of grouping until the categories could not be condensed without losing their individuality. For example, “fine granular surface” and “pitted surface” would not be grouped together because this would not allow an analysis between a granular or pitted cortical surface. Selective coding process After the grouping of observations, the list of categories was considered to be the proposed selective code. Then, using the selective coding system, the notes of the gross observations for each organ were re-read and the observations were placed in one of the categories from the selective code system. Saturation was achieved when all observations fit into a category and no additional categories were needed for completion of the analysis. The same coding process was used for the microscopic observations made by the pathologist. The most common selective codes for the gross and microscopic evaluations are presented in Tables 1 and 2. Table 1. Most prevalent kidney gross observations Observation Prevalence Reddish isolated spots/area 56.10% Granular surface 53.70% Outer cysts 41.50% Lobular in shape 39% Spider-like arteries 36.60% Red discoloration of cortex 36.60% Congested parenchyma 31.70% Red spots in cortex 26.80% Thin cortex 26.80% Brown black discoloration / spots 19.50% Table 2. Most prevalent kidney microscopic observations Observation Prevalence Glomerulosclerosis 29.30% Arteriolosclerosis 24.40% Autolysis 19.50% Acute tubular necrosis 14.60% Chronic interstitial nephritis 14.60% Tubulointerstitial chronic nephritis 12.20% Coarse calcifications 9.76% Interstitial fibrosis 4.88% Metastatic carcinoma 4.88% Nodular glomerulosclerosis 4.88% Thyroidization of tubules 4.88% Statistical analysis A Phi correlation (φ) analysis was conducted among all categories in the selective coding systems (gross and histologic) to determine statistical significance of correlations between the categories. This statistical analysis was conducted using the Graphpad Prism 6 software program. All correlations with a p value less than 0.05 were considered significant. continued on next page A New Resource for Integrated Anatomy Teaching: The Cadaver’s Kidney PG (Pathology Guide) 34 • HAPS Educator Journal of the Human Anatomy and Physiology Society December 2017 Winter Edition Development of the cadaver pathology guide for the kidney After statistically analyzing all of the correlations between the gross and microscopic observations, a list of significant correlations was generated. The list was then analyzed based on basic pathologic principles to determine whether any of the correlations were indirect, or artificially created by the categorization process; those items were removed from the results. The remaining correlations were considered of importance for creating the pathology guide (Table 3). Table 3. Significant correlations between gross and histologic observations Table 3: This table lists the significant correlations between the gross and histologic observations. The phi coefficient is used to detect an association between two variables. All of the significant correlations were positive, meaning that if one variable was