Hardy, J.R., Rees, E., Ling, J., Burman, R., Feuer, D., Broadley, K., & Stone, P. (2001). A prospective survey of the use of dexamethasone on a palliative care unit. Palliative Medicine, 15, 3–8.

To survey the use of dexamethasone on the palliative care wards at this cancer center when prescribed according to the guidelines, to document both the indication for use and any benefit obtained, and to document all side effects incurred.

Consecutive patients started on dexamethasone between April and December 1996 were entered into the survey. The department’s corticosteroid-prescribing policy is as follows: start at moderate dose (8-12 mg per day), wean rapidly to lowest effective dose, monitor closely, prescribe prophylactic nystatin (1 ml four times per day), prescribe prophylactic gastric protectants to all patients with a history of PUD or taking NSAIDs; if no benefit, then discontinue.

• N =106 patients enrolled; by week 6, only 24 patients remained. Only 13 (12.26%) had dyspnea. 
• AGE: Not stated
• MALES (%): Not stated 
• FEMALES (%): Not stated
• CURRENT TREATMENT: Not applicable, other
• KEY DISEASE CHARACTERISTICS: All patients had advanced malignant disease, median survival date of all patients from date of commencement of dexamethasone was 40.5 days.
• OTHER KEY SAMPLE CHARACTERISTICS: Mot common indication for steroids was spinal cord compression (6%), while anorexia (19%), nausea (12%), and low mood (12%) were the most common nonspecific indications. The most common reason for stopping steroids was death or deteriorating condition (48%). Seven patients were taking steroids by the end of eight weeks.

• SITE: Single site   
• SETTING TYPE: Inpatient and outpatient 
• LOCATION: England-Royal Marsden Trust hospitals (London and Surrey)

• PHASE OF CARE: End-of-life care
• APPLICATIONS: Palliative care

Prospective survey

A performa was used to record the reason for starting steroids, the starting and any subsequent doses, the symptoms being palliated, any side effects, and the reasons for stopping steroids. The performa was updated weekly by the ward doctors or research nurses for a maximum of eight weeks. Following discharge, patients were followed-up in the outpatient clinic or at home by telephone. Symptoms were rated on a four-point scale (ranging from 0 to 3) corresponding to none, mild, moderate, and severe. Symptom responses were subsequently recorded as better (decrease in symptom score), worse (increase in symptom score), or no change compared to baseline. Response over time was shown by documenting the proportion of patients with an improvement in a symptom score from baseline at two specific time points: week 2 and at the last assessment. “Best overall response” relates to the best response documented at any time during the treatment with steroids. Side effects were also documented according to the four-point scale. Specific for dyspnea, 5 of 13 (38.5%) patients reported better, 6 of 13 (46.1%) reported unchanged, and 2 of 13 (15.4%) reported worse.

The most common specific reason for initiating steroids was spinal cord compression (6%), followed by cerebral metastases (4%) and then lymphangitis carcinomatosa (4%). The most common nonspecific indications were anorexia (19%), nausea (12%), and low mood (12%). The median starting dose for specific and non-specific indications was 12 mg and 8 mg respectively. In 96 cases, the median duration use was 21.5 days. The most common reasons for stopping steroids include; death/deterioration (48%), tailed off steroids (16%), and trail of steroids ineffective (9%). The symptoms that appeared to get better with steroids are anorexia (73%), nausea (92%), pain (86%), vomiting (94%), bone pain (100%), and all others (73%). The majority of patients complaining of dyspnea or poor mobility showed no change or worsening of symptoms. The most common side effects were oral candida (23% mild and 11% moderate), bruising/petechiae (16% mild and 10% moderate), and proximal myopathy (10% milk and 13% moderate).

Dexamethasone did not improve the symptom of dyspnea in this study. Thirty-nine percent of patients stated the symptom was better, 46% of patients reported no change, and 15% of patients reported dyspnea worsening with steroid use.

• Small sample (< 30)
• Risk of bias (no control group)
• Risk of bias (no blinding)
• Risk of bias (no random assignment)
• Risk of bias (no appropriate attentional control condition)
• Risk of bias (sample characteristics)
• Unintended interventions or applicable interventions not described that would influence results
• Selective outcomes reporting
• Key sample group differences that could influence results
• Measurement validity/reliability questionable 
• Findings not generalizable
• Questionable protocol fidelity
• Subject withdrawals ≥ 10%
• Other limitations/explanation: This study simply reported the findings by reviewing patients in the palliative care program with advanced cancer. No statistical analysis was used to determine if the intervention was successful or not. There is no demographic information regarding the participants other than they all had advanced cancer (no breakdown of age, male or female). No notation of concomitant intervention, no baseline dyspnea statistics (example, how many patients were severe at baseline?). Unknown if more groups were more likely to respond (example, were severe dyspnea patients more likely to respond than mild?). Comorbidities were not identified. Steroid choice and dose not controlled.

If nurses administer dexamethasone, it is imperative they assess for the s/s of oral candida. Nursing staff needs educated on the importance of nystatin as a prophylactic for oral candida. Nursing should be aware of other potential side effects from dexamethasone such as bruising and proximal myopathy.