The investigators anticipate this research will shed light on metric properties of outcome measures of pain and fatigue and the effects of the combined therapy for stroke patients. The overall findings will contribute to guide the choice of proper assessment tool and the development of effective rehabilitation programs.

Hemiplegic shoulder pain affects from 16%-84% of stroke patients and fatigue occurs among 39%-72% of stroke victims, which can have an adverse impact on rehabilitation outcomes by interrupting treatment process, daily functions, and quality of life (QOL) for stroke patients. Considering the high prevalence and detrimental effects of poststroke shoulder pain and fatigue, it is vital for monitoring these symptoms and therapeutic interventions to be developed for those suffering from poststroke pain and fatigue. Therefore, to validate existing assessments of pain/fatigue, and to evaluate the efficacy of potential pain/fatigue interventions are needed.

A vertical Numerical Rating Scale with Faces Rating Scale (NRS-FRS) could be used by all stroke patients as pain and fatigue measurement. The Brief Pain Inventory (BPI) and Brief Fatigue Inventory (BFI) measure pain/fatigue intensity and pain/fatigue interference, which might provide more information regarding pain/fatigue symptoms. The first aim of this research is to examine the psychometric properties of the vertical NRS-FRS, BPI, and BFI.

Functional electrical stimulation (FES) appears to offer benefits in relieving painful hemiplegic shoulder. Previous studies have shown that therapeutic electrical stimulation of the posterior deltoid and supraspinatus muscles significantly reduces shoulder subluxation, but inconsistent results on pain reduction, pain-free range of motion of the affected shoulder joint, and arm function. Conventional rehabilitation usually applies transcutaneous electrical nerve stimulation (TENS) to relieve pain. Bilateral arm training (BAT) is one of the novel interventions in stroke rehabilitation to improve motor recovery of the affected arm, especially proximal part of the upper-limb.Therefore, the second aim of this research is to investigate whether the combined therapy of FES and BAT (FES-BAT) could reduce hemiplegic shoulder pain and improve motor function of the affected arm as compared with TENS plus BAT (TENS-BAT). Whether interventions could further lead to better functional performance and higher QOL would be of interest.

This 3-year project is implemented in three stages and recruit 80 patients with chronic stroke. At the first stage, investigators conduct a test-retest study to investigate the reliability and validity of outcome measures on hemiplegic shoulder pain and poststroke fatigue in 80 patients. At the second stage, investigators conduct a randomized controlled trial (RCT) to investigate the relative effects of the FES-BAT versus TENS-BAT for 40 stroke patients with hemiplegic shoulder pain and moderate-to-mild motor impairment. Treatment regimens are designed to ensure that patients in the 2 groups receive an equivalent intensity of treatment (1 hour/day, 3 days/week for 4 weeks). At the third stage, investigators will determine clinimetric properties of the primary outcome measures, such as responnsiveness, minimal detectable change, and minimal clinical important change.

Inclusion Criteria:

1. first-ever stroke with onset duration more than 3 months;
2. self-reported at least mild intensity of hemiplegic shoulder pain and poststroke fatigue in the past 7 days (the NRS score ≥ 2);
3. Brunnstrom stage III or above and an initial Fugl-Meyer Assessment score of 56 to 79 or 65-100 indicating moderate or mild movement impairment for Stage 2;
4. no serious cognitive deficits (a score of more than 22 on the Mini Mental State Exam);
5. no participation in any experimental rehabilitation or drug studies during the study period; and (6) willing to provide written informed consent prior to study entry.

Exclusion Criteria:

1. history of ventricular arrhythmias, or with a cardiac pacemaker, especially those with cardiac failure with conduction problems;
2. previous contralateral stroke with persistent neurological deficit;
3. a shoulder pathology not related to the stroke (tumor, infection, scapular instability, winged scapula);
4. complicated regional pain syndrome or brachial plexus lesion;
5. diagnosis of epilepsy with history of recurring seizures in the past six months;
6. under irregular analgesia or other fatigue-relieving treatment during the study period because we expect that pain/fatigue medications might complicate measurements of shoulder pain and fatigue. If patients are on regular treatment, it has been so for more than one week;
7. acute pain after operation; and
8. auditory, visual, physical, or mental disabilities that would interfere with patients' ability to comprehend instructions for completing the outcome measures.